wwelsh/Research-Digest
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
Files
main
Research-Digest/tiktok_feed.html
wwelsh 9c4ee28270 initial commit
2025-11-05 12:35:09 -05:00

1305 lines
54 KiB
HTML
Raw Permalink Blame History

<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8" />
<meta name="viewport" content="width=device-width, initial-scale=1.0, user-scalable=no"/>
<title>Research Feed • November 05, 2025</title>
<style>
* {
box-sizing: border-box;
margin: 0;
padding: 0;
}
:root {
--bg: #000000;
--text: #ffffff;
--muted: #a0a0a0;
--card-bg: #1a1a1a;
--border: #2a2a2a;
--accent: #ff6b6b;
--heart-red: #ff4458;
--layman-bg: #1f2937;
--layman-border: #60a5fa;
}
body {
font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, sans-serif;
background: var(--bg);
color: var(--text);
overflow-x: hidden;
-webkit-font-smoothing: antialiased;
}
#feed-container {
height: 100vh;
overflow-y: scroll;
scroll-snap-type: y mandatory;
-webkit-overflow-scrolling: touch;
padding-top: 60px; /* Space for fixed header */
}
.paper-card {
min-height: 100vh;
scroll-snap-align: start;
scroll-snap-stop: always;
display: flex;
flex-direction: column;
justify-content: center;
padding: 2rem 1.5rem;
position: relative;
border-bottom: 1px solid var(--border);
}
.interest-badge {
display: inline-block;
background: var(--accent);
color: white;
padding: 0.4rem 0.9rem;
border-radius: 20px;
font-size: 0.7rem;
font-weight: 700;
text-transform: uppercase;
letter-spacing: 0.5px;
margin-bottom: 1rem;
}
.difficulty-badge {
display: inline-block;
padding: 0.3rem 0.7rem;
border-radius: 15px;
font-size: 0.7rem;
font-weight: 600;
margin-left: 0.5rem;
}
.paper-title {
font-size: 1.5rem;
font-weight: 800;
line-height: 1.3;
margin-bottom: 1rem;
color: var(--text);
}
.layman-box {
background: var(--layman-bg);
border-left: 3px solid var(--layman-border);
padding: 1rem;
margin-bottom: 1rem;
border-radius: 8px;
font-size: 0.95rem;
line-height: 1.6;
color: #94a3b8;
font-style: italic;
}
.summary {
color: var(--muted);
font-size: 0.95rem;
line-height: 1.6;
margin-bottom: 1.5rem;
}
.paper-meta {
display: flex;
justify-content: space-between;
align-items: center;
margin-bottom: 1rem;
color: var(--muted);
font-size: 0.8rem;
}
.category-tag {
background: #1e3a5f;
color: #60a5fa;
padding: 0.3rem 0.7rem;
border-radius: 15px;
font-size: 0.75rem;
font-weight: 600;
}
.links {
display: flex;
gap: 1.5rem;
margin-bottom: 2rem;
}
.links a {
color: #6ba3ff;
text-decoration: none;
font-size: 0.9rem;
font-weight: 600;
display: flex;
align-items: center;
gap: 0.3rem;
}
.links a:active {
color: var(--accent);
}
/* Fixed header with export button */
.fixed-header {
position: fixed;
top: 0;
left: 0;
right: 0;
background: rgba(0, 0, 0, 0.95);
backdrop-filter: blur(10px);
padding: 0.8rem 1.5rem;
display: flex;
align-items: center;
justify-content: space-between;
border-bottom: 1px solid var(--border);
z-index: 200;
}
.like-counter {
display: flex;
align-items: center;
gap: 0.5rem;
color: var(--muted);
font-size: 0.9rem;
}
.export-button {
background: linear-gradient(135deg, var(--accent), #ffa94d);
color: white;
border: none;
padding: 0.6rem 1.2rem;
border-radius: 20px;
font-weight: 700;
font-size: 0.85rem;
cursor: pointer;
transition: all 0.2s ease;
-webkit-tap-highlight-color: transparent;
opacity: 0.5;
pointer-events: none;
}
.export-button.active {
opacity: 1;
pointer-events: auto;
}
.export-button.active:active {
transform: scale(0.95);
}
.like-button {
position: fixed;
bottom: 2rem;
right: 1.5rem;
width: 60px;
height: 60px;
border-radius: 50%;
background: rgba(26, 26, 26, 0.9);
border: 2px solid var(--border);
display: flex;
align-items: center;
justify-content: center;
font-size: 1.8rem;
cursor: pointer;
transition: all 0.2s ease;
z-index: 100;
-webkit-tap-highlight-color: transparent;
}
.like-button:active {
transform: scale(0.9);
}
.like-button.liked {
background: var(--heart-red);
border-color: var(--heart-red);
animation: heartbeat 0.3s ease;
}
@keyframes heartbeat {
0%, 100% { transform: scale(1); }
50% { transform: scale(1.2); }
}
.scroll-indicator {
position: fixed;
bottom: 1rem;
left: 50%;
transform: translateX(-50%);
color: var(--muted);
font-size: 0.8rem;
animation: bounce 2s infinite;
z-index: 50;
}
@keyframes bounce {
0%, 100% { transform: translateX(-50%) translateY(0); }
50% { transform: translateX(-50%) translateY(-10px); }
}
.hide-indicator {
display: none;
}
</style>
</head>
<body>
<!-- Fixed Header with Export Button -->
<div class="fixed-header">
<div class="like-counter">
<span>♥</span>
<span><span id="likeCount">0</span> liked</span>
</div>
<button class="export-button" id="exportButton">Export Likes</button>
</div>
<div id="feed-container"></div>
<div class="like-button" id="likeButton">
<span id="heartIcon">♡</span>
</div>
<div class="scroll-indicator" id="scrollIndicator">
↓ Scroll to explore
</div>
<script>
// ============================================
// EMBEDDED PAPERS DATA
// ============================================
const papers = [
{
"title": "Nesterov-Accelerated Robust Federated Learning Over Byzantine Adversaries",
"summary": " We investigate robust federated learning, where a group of workers collaboratively train a shared model under the orchestration of a central server in the presence of Byzantine adversaries . We propose a Byzantine-resilient Nesterov-Accelerated Federated Learning (Byrd-NAFL) algorithm .",
"link": "http://arxiv.org/abs/2511.02657v1",
"pdf_link": "https://arxiv.org/pdf/2511.02657.pdf",
"arxiv_id": "2511.02657",
"category": "cs.LG",
"published": "2025-11-04",
"relevance_score": 3,
"matched_keywords": [
"accelerat"
],
"difficulty": "🟢 Applied",
"layman": "This research enhances privacy-preserving AI.",
"interest_category": "Efficient ML / Edge AI"
},
{
"title": "DANIEL: A Distributed and Scalable Approach for Global Representation Learning with EHR Applications",
"summary": " Classical probabilistic graphical models face challenges in modern data environments . We revisit the Ising model and develop a distributed framework that enables scalable representation learning from large-scale binary data . We evaluate our algorithm on multi-institutional electronic health record (EHR) datasets from 58,248 patients .",
"link": "http://arxiv.org/abs/2511.02754v1",
"pdf_link": "https://arxiv.org/pdf/2511.02754.pdf",
"arxiv_id": "2511.02754",
"category": "stat.ME",
"published": "2025-11-04",
"relevance_score": 2,
"matched_keywords": [
"privacy",
"federated"
],
"difficulty": "🟢 Applied",
"layman": "This research explores techniques in privacy-preserving AI.",
"interest_category": "Privacy-Preserving ML"
},
{
"title": "Adaptive Neighborhood-Constrained Q Learning for Offline Reinforcement Learning",
"summary": " Offline reinforcement learning (RL) suffers from extrapolation errors induced by out-of-distribution (OOD) actions . To overcome these limitations, we propose a new neighborhood constraint that restricts action selection in the Bellman target to the union of neighborhoods of dataset actions . Theoretically, the constraint bounds extrapolation error and distribution shift under certain conditions, but also approximates the support constraint without requiring behavior policy modeling .",
"link": "http://arxiv.org/abs/2511.02567v1",
"pdf_link": "https://arxiv.org/pdf/2511.02567.pdf",
"arxiv_id": "2511.02567",
"category": "cs.LG",
"published": "2025-11-04",
"relevance_score": 2,
"matched_keywords": [
"art",
"design"
],
"difficulty": "🟢 Applied",
"layman": "This research tackles the problem of language AI.",
"interest_category": "Creative AI / Emotion"
},
{
"title": "EPARA: Parallelizing Categorized AI Inference in Edge Clouds",
"summary": " EPARA is an end-to-end AI parallel inference framework aimed at enhancing the edge AI serving capability . EPARA consists of a task-categorized parallelism allocator that decides the parallel mode of each task, a distributed request handler that performs the calculation for the specific request, and a state-aware scheduler that periodically updates service placement in edge clouds .",
"link": "http://arxiv.org/abs/2511.00603v1",
"pdf_link": "https://arxiv.org/pdf/2511.00603.pdf",
"arxiv_id": "2511.00603",
"category": "cs.DC",
"published": "2025-11-01",
"relevance_score": 8,
"matched_keywords": [
"embedded",
"edge",
"resource",
"device"
],
"difficulty": "🟢 Applied",
"layman": "This research enhances language AI.",
"interest_category": "Lightweight Systems"
},
{
"title": "Forecasting Future Anatomies: Longitudianl Brain Mri-to-Mri Prediction",
"summary": " Neuroimaging has important implications for studying neurodegenerative diseases such as Alzheimer's disease . Most existing approaches predict future cognitive scores or clinical outcomes . Here we investigate longitudinal MRI image-to-image prediction that forecasts a participant's entire brain MRI several years into the future .",
"link": "http://arxiv.org/abs/2511.02558v1",
"pdf_link": "https://arxiv.org/pdf/2511.02558.pdf",
"arxiv_id": "2511.02558",
"category": "cs.CV",
"published": "2025-11-04",
"relevance_score": 1,
"matched_keywords": [
"local"
],
"difficulty": "🟢 Applied",
"layman": "This research creating new content with computer vision.",
"interest_category": "Offline-First / Local AI"
},
{
"title": "Curriculum Design for Trajectory-Constrained Agent: Compressing Chain-of-Thought Tokens in LLMs",
"summary": " Training agents to operate under strict constraints during deployment presents significant challenges . In this work, we propose a curriculum learning strategy that gradually tightens constraints during training, enabling the agent to incrementally master the deployment requirements . We provide theoretical analysis using an RL agent in a binary-tree Markov Decision Process .",
"link": "http://arxiv.org/abs/2511.02690v1",
"pdf_link": "https://arxiv.org/pdf/2511.02690.pdf",
"arxiv_id": "2511.02690",
"category": "cs.LG",
"published": "2025-11-04",
"relevance_score": 4,
"matched_keywords": [
"compression",
"inference",
"accelerat"
],
"difficulty": "🟢 Applied",
"layman": "This research proposes a method for language AI.",
"interest_category": "Efficient ML / Edge AI"
},
{
"title": "Orion-MSP: Multi-Scale Sparse Attention for Tabular In-Context Learning",
"summary": " Tabular data remain the predominant format for real-world applications . Developing effective neural models for tabular data remains challenging due to heterogeneous feature types and complex interactions occurring at multiple scales . Orion-MSP is a tabular ICL architecture featuring three key innovations: multi-scale processing to capture hierarchical feature interactions .",
"link": "http://arxiv.org/abs/2511.02818v1",
"pdf_link": "https://arxiv.org/pdf/2511.02818.pdf",
"arxiv_id": "2511.02818",
"category": "cs.AI",
"published": "2025-11-04",
"relevance_score": 0,
"matched_keywords": [],
"difficulty": "🟢 Applied",
"layman": "This research explores techniques in machine learning.",
"interest_category": "Privacy-Preserving ML"
},
{
"title": "STAR-VAE: Latent Variable Transformers for Scalable and Controllable Molecular Generation",
"summary": " STAR-VAE (Selfies-encoded, Transformer-based, AutoRegressive Variational Auto Encoder) is a scalable latent-variable framework . It is trained on 79 million drug-like molecules from PubChem, using SELFIES to guarantee syntactic validity .",
"link": "http://arxiv.org/abs/2511.02769v1",
"pdf_link": "https://arxiv.org/pdf/2511.02769.pdf",
"arxiv_id": "2511.02769",
"category": "cs.LG",
"published": "2025-11-04",
"relevance_score": 2,
"matched_keywords": [
"generative",
"art"
],
"difficulty": "🟢 Applied",
"layman": "This research presents techniques for machine learning.",
"interest_category": "Creative AI / Emotion"
},
{
"title": "Real-time Continual Learning on Intel Loihi 2",
"summary": " AI systems on edge devices face a critical challenge in open-world environments . Online continual learning (OCL) remains challenging in power-constrained settings . We present a neuromorphic solution called CLP-SNN: a spiking neural network architecture for Continually Learning Prototypes .",
"link": "http://arxiv.org/abs/2511.01553v1",
"pdf_link": "https://arxiv.org/pdf/2511.01553.pdf",
"arxiv_id": "2511.01553",
"category": "cs.LG",
"published": "2025-11-03",
"relevance_score": 6,
"matched_keywords": [
"iot",
"edge",
"constrained",
"device"
],
"difficulty": "🟢 Applied",
"layman": "This research introduces a new approach to edge computing.",
"interest_category": "Lightweight Systems"
},
{
"title": "In Situ Training of Implicit Neural Compressors for Scientific Simulations via Sketch-Based Regularization",
"summary": " Focusing on implicit neural representations, we present a novel in situ training protocol that employs limited memory buffers of full and sketched data samples . We evaluate our method on a variety of complex simulation data in two and three dimensions .",
"link": "http://arxiv.org/abs/2511.02659v1",
"pdf_link": "https://arxiv.org/pdf/2511.02659.pdf",
"arxiv_id": "2511.02659",
"category": "cs.LG",
"published": "2025-11-04",
"relevance_score": 1,
"matched_keywords": [
"offline"
],
"difficulty": "🟡 Advanced",
"layman": "This research introduces a new approach to machine learning.",
"interest_category": "Offline-First / Local AI"
},
{
"title": "Optimal Singular Damage: Efficient LLM Inference in Low Storage Regimes",
"summary": " Large language models (LLMs) are increasingly prevalent across diverse applications . However, their enormous size limits storage and processing capabilities to a few well-resourced stakeholders . This paper focuses on efficient storage of parameter updates in pre-trained models after fine-tuning .",
"link": "http://arxiv.org/abs/2511.02681v1",
"pdf_link": "https://arxiv.org/pdf/2511.02681.pdf",
"arxiv_id": "2511.02681",
"category": "cs.CL",
"published": "2025-11-04",
"relevance_score": 6,
"matched_keywords": [
"efficient",
"inference"
],
"difficulty": "🟢 Applied",
"layman": "This research explores techniques in language AI.",
"interest_category": "Efficient ML / Edge AI"
},
{
"title": "Agent-Omni: Test-Time Multimodal Reasoning via Model Coordination for Understanding Anything",
"summary": " Multimodal large language models have shown strong capabilities but remain limited to fixed modality pairs . Building fully omni-capable models that can integrate text, images, audio, and video remains impractical and lacks robust reasoning support . We propose an Agent-Omni framework that coordinates existing foundation models through a master-agent system .",
"link": "http://arxiv.org/abs/2511.02834v1",
"pdf_link": "https://arxiv.org/pdf/2511.02834.pdf",
"arxiv_id": "2511.02834",
"category": "cs.AI",
"published": "2025-11-04",
"relevance_score": 0,
"matched_keywords": [],
"difficulty": "🟢 Applied",
"layman": "This research explores techniques in language AI.",
"interest_category": "Privacy-Preserving ML"
},
{
"title": "When One Modality Sabotages the Others: A Diagnostic Lens on Multimodal Reasoning",
"summary": " In this paper, we introduce modality sabotage, a diagnostic failure mode in which a high-confidence unimodal error overrides other evidence and misleads the fused result . Applying our diagnostic layer in a case study on multimodal emotion recognition benchmarks with foundation models revealed systematic reliability profiles .",
"link": "http://arxiv.org/abs/2511.02794v1",
"pdf_link": "https://arxiv.org/pdf/2511.02794.pdf",
"arxiv_id": "2511.02794",
"category": "cs.AI",
"published": "2025-11-04",
"relevance_score": 2,
"matched_keywords": [
"emotion",
"art"
],
"difficulty": "🟢 Applied",
"layman": "This research forecasting language AI.",
"interest_category": "Creative AI / Emotion"
},
{
"title": "Boosting performance of computer vision applications through embedded GPUs on the edge",
"summary": " Edge computing can be used to offload certain high intensive tasks . Edge computing is usually composed of devices with limited capacity, which may impact in users quality of experience . This work proposes the use of embedded devices with graphics processing units (GPUs) to overcome such limitation .",
"link": "http://arxiv.org/abs/2511.01129v1",
"pdf_link": "https://arxiv.org/pdf/2511.01129.pdf",
"arxiv_id": "2511.01129",
"category": "cs.CV",
"published": "2025-11-03",
"relevance_score": 10,
"matched_keywords": [
"embedded",
"edge",
"resource",
"device"
],
"difficulty": "🟢 Applied",
"layman": "This research presents techniques for computer vision.",
"interest_category": "Lightweight Systems"
},
{
"title": "Agentic AI for Mobile Network RAN Management and Optimization",
"summary": " Agentic AI represents a new paradigm for automating complex systems by using Large AI Models (LAMs) to provide human-level cognitive abilities with multimodal perception, planning, memory, and reasoning capabilities . The complexity of 5G and upcoming 6G networks renders manual optimization ineffective . Despite its rapid advances, there is no established framework outlining the foundational components and operational principles .",
"link": "http://arxiv.org/abs/2511.02532v1",
"pdf_link": "https://arxiv.org/pdf/2511.02532.pdf",
"arxiv_id": "2511.02532",
"category": "cs.AI",
"published": "2025-11-04",
"relevance_score": 3,
"matched_keywords": [
"mobile"
],
"difficulty": "🟢 Applied",
"layman": "This research presents techniques for machine learning.",
"interest_category": "Offline-First / Local AI"
},
{
"title": "Accelerated Frank-Wolfe Algorithms: Complementarity Conditions and Sparsity",
"summary": " We develop new algorithms in Frank-Wolfe family for minimizing smooth convex functions over compact convex sets . A key technical ingredient is a complementarity condition that captures solution sparsity -- face dimension for polytopes and rank for matrices . Our results close a gap on how to accelerate recent advancements in linearly-converging FW algorithms .",
"link": "http://arxiv.org/abs/2511.02821v1",
"pdf_link": "https://arxiv.org/pdf/2511.02821.pdf",
"arxiv_id": "2511.02821",
"category": "math.OC",
"published": "2025-11-04",
"relevance_score": 3,
"matched_keywords": [
"accelerat"
],
"difficulty": "🟡 Advanced",
"layman": "This research speeds up machine learning.",
"interest_category": "Efficient ML / Edge AI"
},
{
"title": "Bringing Private Reads to Hyperledger Fabric via Private Information Retrieval",
"summary": " The prototype achieves an average end-to-end latency of 113 ms and a peer-side execution time below 42 ms . Storage profiling across three channel configurations shows near-linear growth .",
"link": "http://arxiv.org/abs/2511.02656v1",
"pdf_link": "https://arxiv.org/pdf/2511.02656.pdf",
"arxiv_id": "2511.02656",
"category": "cs.CR",
"published": "2025-11-04",
"relevance_score": 6,
"matched_keywords": [
"privacy",
"encrypted",
"private"
],
"difficulty": "🟢 Applied",
"layman": "This research proposes a method for privacy-preserving AI.",
"interest_category": "Privacy-Preserving ML"
},
{
"title": "Emotional Contagion in Code: How GitHub Emoji Reactions Shape Developer Collaboration",
"summary": " Developer communities increasingly rely on emoji reactions to communicate, but we know little about how these signals spread and influence technical discussions . We analyzed 2,098 GitHub issues and pull requests across 50 popular repositories to understand emotional contagion in software development . Our findings reveal a surprisingly positive emotional landscape: 57.4% of discussions carry positive sentiment .",
"link": "http://arxiv.org/abs/2511.02515v1",
"pdf_link": "https://arxiv.org/pdf/2511.02515.pdf",
"arxiv_id": "2511.02515",
"category": "cs.HC",
"published": "2025-11-04",
"relevance_score": 4,
"matched_keywords": [
"emotion",
"sentiment"
],
"difficulty": "🟢 Applied",
"layman": "This research understanding emotions in emotion AI.",
"interest_category": "Creative AI / Emotion"
},
{
"title": "Neuro-Inspired Task Offloading in Edge-IoT Networks Using Spiking Neural Networks",
"summary": " Traditional task offloading strategies in edge computing often rely on static heuristics or data-intensive machine learning models . We propose a novel task-offloading framework based on Spiking Neural Networks inspired by the efficiency and adaptability of biological neural systems .",
"link": "http://arxiv.org/abs/2511.01127v1",
"pdf_link": "https://arxiv.org/pdf/2511.01127.pdf",
"arxiv_id": "2511.01127",
"category": "cs.DC",
"published": "2025-11-03",
"relevance_score": 10,
"matched_keywords": [
"iot",
"edge",
"resource",
"constrained"
],
"difficulty": "🟢 Applied",
"layman": "This research introduces a new approach to edge computing.",
"interest_category": "Lightweight Systems"
},
{
"title": "Agentic World Modeling for 6G: Near-Real-Time Generative State-Space Reasoning",
"summary": " We reframe open radio access network (O-RAN) near-real-time (Near-RT) control via counterfactual dynamics and a world modeling (WM) paradigm . This enables quantitative \"what-if\" forecasting beyond large language models as the primary modeling primitive . Actions such as physical resource blocks are treated as first-class control inputs in a causal world model .",
"link": "http://arxiv.org/abs/2511.02748v1",
"pdf_link": "https://arxiv.org/pdf/2511.02748.pdf",
"arxiv_id": "2511.02748",
"category": "cs.NI",
"published": "2025-11-04",
"relevance_score": 1,
"matched_keywords": [
"offline"
],
"difficulty": "🟢 Applied",
"layman": "This research forecasting language AI.",
"interest_category": "Offline-First / Local AI"
},
{
"title": "In Good GRACEs: Principled Teacher Selection for Knowledge Distillation",
"summary": " Knowledge distillation is an efficient strategy to use data generated by large \"teacher\" language models to train smaller capable \"student\" models . But selecting the optimal teacher for a specific student-task combination requires expensive trial-and-error . We propose a lightweight score called GRACE to quantify how effective a teacher will be for post-training a student model .",
"link": "http://arxiv.org/abs/2511.02833v1",
"pdf_link": "https://arxiv.org/pdf/2511.02833.pdf",
"arxiv_id": "2511.02833",
"category": "cs.LG",
"published": "2025-11-04",
"relevance_score": 10,
"matched_keywords": [
"efficient",
"edge",
"distillation",
"lightweight"
],
"difficulty": "🟡 Advanced",
"layman": "This research makes more efficient language AI.",
"interest_category": "Efficient ML / Edge AI"
},
{
"title": "Fast, Private, and Protected: Safeguarding Data Privacy and Defending Against Model Poisoning Attacks in Federated Learning",
"summary": " Federated Learning (FL) is a distributed training paradigm wherein participants collaborate to build a global model while ensuring the privacy of the involved data . FPP employs a reputation-based mechanism to mitigate the participation of attackers .",
"link": "http://arxiv.org/abs/2511.02797v1",
"pdf_link": "https://arxiv.org/pdf/2511.02797.pdf",
"arxiv_id": "2511.02797",
"category": "cs.LG",
"published": "2025-11-04",
"relevance_score": 12,
"matched_keywords": [
"privacy",
"federated",
"secure",
"private"
],
"difficulty": "🟢 Applied",
"layman": "This research protecting data privacy in privacy-preserving AI.",
"interest_category": "Privacy-Preserving ML"
},
{
"title": "Audience Amplified: Virtual Audiences in Asynchronously Performed AR Theater",
"summary": " Users can wander around naturally and engage in AR theater with virtual audiences trained from real audiences using imitation learning . Virtual dancers emerge from the surroundings, accompanied by a digitally simulated audience .",
"link": "http://arxiv.org/abs/2511.02807v1",
"pdf_link": "https://arxiv.org/pdf/2511.02807.pdf",
"arxiv_id": "2511.02807",
"category": "cs.HC",
"published": "2025-11-04",
"relevance_score": 4,
"matched_keywords": [
"sentiment",
"art",
"audio"
],
"difficulty": "🟢 Applied",
"layman": "This research enhances speech processing.",
"interest_category": "Creative AI / Emotion"
},
{
"title": "Split Learning-Enabled Framework for Secure and Light-weight Internet of Medical Things Systems",
"summary": " Conventional deep learning methods are impractical due to resource limitations . Federated Learning (FL) suffers from high communication overhead and vulnerability to non-IID data . Split learning (SL) based framework for IoT malware detection through image-based classification .",
"link": "http://arxiv.org/abs/2511.00336v1",
"pdf_link": "https://arxiv.org/pdf/2511.00336.pdf",
"arxiv_id": "2511.00336",
"category": "cs.CR",
"published": "2025-11-01",
"relevance_score": 8,
"matched_keywords": [
"iot",
"edge",
"resource",
"constrained",
"device"
],
"difficulty": "🟢 Applied",
"layman": "This research distributed machine learning across computer vision.",
"interest_category": "Lightweight Systems"
},
{
"title": "RL-Aided Cognitive ISAC: Robust Detection and Sensing-Communication Trade-offs",
"summary": " This paper proposes a reinforcement learning (RL)-aided cognitive framework for massive MIMO-based integrated sensing and communication (ISAC) systems . The focus is on enhancing radar sensing performance in environments with unknown and dynamic disturbance characteristics .",
"link": "http://arxiv.org/abs/2511.02672v1",
"pdf_link": "https://arxiv.org/pdf/2511.02672.pdf",
"arxiv_id": "2511.02672",
"category": "eess.SP",
"published": "2025-11-04",
"relevance_score": 1,
"matched_keywords": [
"edge"
],
"difficulty": "🟢 Applied",
"layman": "This research enhances edge computing.",
"interest_category": "Offline-First / Local AI"
},
{
"title": "Apriel-H1: Towards Efficient Enterprise Reasoning Models",
"summary": " Hybrid LLMs combine transformer attention and SSM sequence mixers for efficient reasoning at 15B model size . High inference throughput is critical for agentic tasks, long-context reasoning, efficient deployment under high request loads, and more efficient test-time compute scaling .",
"link": "http://arxiv.org/abs/2511.02651v1",
"pdf_link": "https://arxiv.org/pdf/2511.02651.pdf",
"arxiv_id": "2511.02651",
"category": "cs.LG",
"published": "2025-11-04",
"relevance_score": 6,
"matched_keywords": [
"efficient",
"distillation",
"inference"
],
"difficulty": "🟢 Applied",
"layman": "This research achieves better language AI.",
"interest_category": "Efficient ML / Edge AI"
},
{
"title": "Enhancing Federated Learning Privacy with QUBO",
"summary": " Federated learning (FL) is a widely used method for training machine learning (ML) models in a scalable way while preserving privacy . The risk of exposing sensitive data increases cumulatively as the number of iterations where a client's updates are included in the aggregated model increases . Attackers can launch membership inference attacks (MIA), property inference attacks, and model inversion attacks .",
"link": "http://arxiv.org/abs/2511.02785v1",
"pdf_link": "https://arxiv.org/pdf/2511.02785.pdf",
"arxiv_id": "2511.02785",
"category": "cs.LG",
"published": "2025-11-04",
"relevance_score": 6,
"matched_keywords": [
"privacy",
"federated"
],
"difficulty": "🟢 Applied",
"layman": "This research protecting data privacy in privacy-preserving AI.",
"interest_category": "Privacy-Preserving ML"
},
{
"title": "SigmaCollab: An Application-Driven Dataset for Physically Situated Collaboration",
"summary": " We introduce SigmaCollab, a dataset enabling research on physically situated human-AI collaboration . The dataset consists of a set of 85 sessions in which untrained participants were guided by a mixed-reality assistive AI agent in performing procedural tasks in the physical world .",
"link": "http://arxiv.org/abs/2511.02560v1",
"pdf_link": "https://arxiv.org/pdf/2511.02560.pdf",
"arxiv_id": "2511.02560",
"category": "cs.HC",
"published": "2025-11-04",
"relevance_score": 2,
"matched_keywords": [
"art",
"audio"
],
"difficulty": "🟢 Applied",
"layman": "This research explores techniques in speech processing.",
"interest_category": "Creative AI / Emotion"
},
{
"title": "Tetris: An SLA-aware Application Placement Strategy in the Edge-Cloud Continuum",
"summary": " An Edge-Cloud Continuum integrates edge and cloud resources to provide a flexible and scalable infrastructure . This paradigm can minimize latency by processing data closer to the source at the edge while leveraging the vast computational power of the cloud for more intensive tasks .",
"link": "http://arxiv.org/abs/2511.00294v1",
"pdf_link": "https://arxiv.org/pdf/2511.00294.pdf",
"arxiv_id": "2511.00294",
"category": "cs.DC",
"published": "2025-10-31",
"relevance_score": 4,
"matched_keywords": [
"edge",
"resource"
],
"difficulty": "🟢 Applied",
"layman": "This research running AI locally on devices for edge computing.",
"interest_category": "Lightweight Systems"
},
{
"title": "VecComp: Vector Computing via MIMO Digital Over-the-Air Computation",
"summary": " ChannelComp framework has proposed digital over-the-air computation by designing digital modulations that enable the computation of arbitrary functions . This framework is intended for applications that favor local information processing over the mere acquisition of data .",
"link": "http://arxiv.org/abs/2511.02765v1",
"pdf_link": "https://arxiv.org/pdf/2511.02765.pdf",
"arxiv_id": "2511.02765",
"category": "cs.LG",
"published": "2025-11-04",
"relevance_score": 1,
"matched_keywords": [
"local"
],
"difficulty": "🟢 Applied",
"layman": "This research proposes a method for machine learning.",
"interest_category": "Offline-First / Local AI"
},
{
"title": "Can Visual Input Be Compressed? A Visual Token Compression Benchmark for Large Multimodal Models",
"summary": " Large multimodal models (LMMs) often suffer from severe inference inefficiency due to the large number of visual tokens introduced by image encoders . Recent token compression methods, such as pruning and merging, have shown promise in reducing redundancy . In this work, we present UniPruneBench, a unified and extensible benchmark for visual token pruning .",
"link": "http://arxiv.org/abs/2511.02650v1",
"pdf_link": "https://arxiv.org/pdf/2511.02650.pdf",
"arxiv_id": "2511.02650",
"category": "cs.CV",
"published": "2025-11-04",
"relevance_score": 8,
"matched_keywords": [
"efficient",
"compression",
"pruning",
"inference"
],
"difficulty": "🟢 Applied",
"layman": "This research reduces language AI.",
"interest_category": "Efficient ML / Edge AI"
},
{
"title": "GeoCrossBench: Cross-Band Generalization for Remote Sensing",
"summary": " GeoCrossBench is an extension of the popular GeoBench benchmark with a new evaluation protocol . It tests the in-distribution performance; generalization to satellites with no band overlap . ChiViT significantly outperforms the runner-up DINOv3 . Fine-tuning just the last linear layer of these models can get relatively consistent performance across all satellites .",
"link": "http://arxiv.org/abs/2511.02831v1",
"pdf_link": "https://arxiv.org/pdf/2511.02831.pdf",
"arxiv_id": "2511.02831",
"category": "cs.LG",
"published": "2025-11-04",
"relevance_score": 0,
"matched_keywords": [],
"difficulty": "🟢 Applied",
"layman": "This research explores techniques in machine learning.",
"interest_category": "Privacy-Preserving ML"
},
{
"title": "1 PoCo: Agentic Proof-of-Concept Exploit Generation for Smart Contracts",
"summary": " Smart contracts operate in a highly adversarial environment, where vulnerabilities can lead to substantial financial losses . POCO autonomously generates PoC exploits in an agentic manner by interacting with a set of code-execution tools in a Reason-Act-Observe loop .",
"link": "http://arxiv.org/abs/2511.02780v1",
"pdf_link": "https://arxiv.org/pdf/2511.02780.pdf",
"arxiv_id": "2511.02780",
"category": "cs.CR",
"published": "2025-11-04",
"relevance_score": 3,
"matched_keywords": [
"art"
],
"difficulty": "🟢 Applied",
"layman": "This research explores techniques in edge computing.",
"interest_category": "Creative AI / Emotion"
},
{
"title": "Edge AI in Highly Volatile Environments: Is Fairness Worth the Accuracy Trade-off?",
"summary": " Federated learning (FL) has emerged as a transformative paradigm for edge intelligence, enabling collaborative model training while preserving data privacy across distributed personal devices . The inherent volatility of edge environments poses significant challenges for achieving high accuracy and fairness in client participation .",
"link": "http://arxiv.org/abs/2511.01737v1",
"pdf_link": "https://arxiv.org/pdf/2511.01737.pdf",
"arxiv_id": "2511.01737",
"category": "cs.LG",
"published": "2025-11-03",
"relevance_score": 6,
"matched_keywords": [
"edge",
"resource",
"device"
],
"difficulty": "🟢 Applied",
"layman": "This research protecting data privacy in privacy-preserving AI.",
"interest_category": "Lightweight Systems"
},
{
"title": "Dynamic Priors in Bayesian Optimization for Hyperparameter Optimization",
"summary": " Hyperparameter optimization (HPO) based on Bayesian optimization (BO) supports users in designing models well-suited for a given dataset . HPO has proven its effectiveness on several applications, ranging from classical machine learning for tabular data to deep neural networks for computer vision and transformers for natural language processing .",
"link": "http://arxiv.org/abs/2511.02570v1",
"pdf_link": "https://arxiv.org/pdf/2511.02570.pdf",
"arxiv_id": "2511.02570",
"category": "cs.LG",
"published": "2025-11-04",
"relevance_score": 1,
"matched_keywords": [
"edge"
],
"difficulty": "🟢 Applied",
"layman": "This research optimizes computer vision.",
"interest_category": "Offline-First / Local AI"
},
{
"title": "Federated Attention: A Distributed Paradigm for Collaborative LLM Inference over Edge Networks",
"summary": " Large language models (LLMs) are proliferating rapidly at the edge, delivering intelligent capabilities across diverse application scenarios . However, their practical deployment in collaborative scenarios confronts challenges: privacy vulnerabilities, communication overhead, and computational bottlenecks . To address these, we propose Federated Attention (FedAttn), which integrates the federated paradigm into the self-attention mechanism .",
"link": "http://arxiv.org/abs/2511.02647v1",
"pdf_link": "https://arxiv.org/pdf/2511.02647.pdf",
"arxiv_id": "2511.02647",
"category": "cs.DC",
"published": "2025-11-04",
"relevance_score": 6,
"matched_keywords": [
"edge",
"inference"
],
"difficulty": "🟢 Applied",
"layman": "This research protecting data privacy in language AI.",
"interest_category": "Efficient ML / Edge AI"
},
{
"title": "Adam Reduces a Unique Form of Sharpness: Theoretical Insights Near the Minimizer Manifold",
"summary": " Adam implicitly reduces a unique form of sharpness measure shaped by its adaptive updates, leading to qualitatively different solutions from SGD . In solving sparse linear regression with diagonal linear networks, this distinction enables Adam to achieve better sparsity and generalization .",
"link": "http://arxiv.org/abs/2511.02773v1",
"pdf_link": "https://arxiv.org/pdf/2511.02773.pdf",
"arxiv_id": "2511.02773",
"category": "cs.LG",
"published": "2025-11-04",
"relevance_score": 1,
"matched_keywords": [
"differential"
],
"difficulty": "🟡 Advanced",
"layman": "This research reduces machine learning.",
"interest_category": "Privacy-Preserving ML"
},
{
"title": "The Collaboration Gap",
"summary": " The trajectory of AI development suggests we will increasingly rely on agent-based systems composed of independently developed agents . Success of these systems will critically depend on effective collaboration among these heterogeneous agents . Few empirical studies have evaluated such agent-agent collaboration at scale . We propose a collaborative maze-solving benchmark that isolates collaborative capabilities .",
"link": "http://arxiv.org/abs/2511.02687v1",
"pdf_link": "https://arxiv.org/pdf/2511.02687.pdf",
"arxiv_id": "2511.02687",
"category": "cs.AI",
"published": "2025-11-04",
"relevance_score": 2,
"matched_keywords": [
"art",
"design"
],
"difficulty": "🟢 Applied",
"layman": "This research explores techniques in machine learning.",
"interest_category": "Creative AI / Emotion"
},
{
"title": "Energy-Efficient Hardware Acceleration of Whisper ASR on a CGLA",
"summary": " Whisper's core computational kernel was evaluated on a general-purpose Coarse-Grained Linear Arrays (CGLAs) accelerator . Whisper is 1.90x more efficient than the NVIDIA Jetson AGX Orin .",
"link": "http://arxiv.org/abs/2511.02269v1",
"pdf_link": "https://arxiv.org/pdf/2511.02269.pdf",
"arxiv_id": "2511.02269",
"category": "cs.AR",
"published": "2025-11-04",
"relevance_score": 4,
"matched_keywords": [
"edge",
"constrained",
"device"
],
"difficulty": "🟢 Applied",
"layman": "This research tackles the problem of speech processing.",
"interest_category": "Lightweight Systems"
},
{
"title": "Redundancy Maximization as a Principle of Associative Memory Learning",
"summary": " Associative memory, traditionally modeled by Hopfield networks, enables the retrieval of previously stored patterns from partial or noisy cues . We use redundancy as an information-theoretic learning goal, which is directly optimized for each neuron .",
"link": "http://arxiv.org/abs/2511.02584v1",
"pdf_link": "https://arxiv.org/pdf/2511.02584.pdf",
"arxiv_id": "2511.02584",
"category": "cs.IT",
"published": "2025-11-04",
"relevance_score": 1,
"matched_keywords": [
"local"
],
"difficulty": "🟢 Applied",
"layman": "This research explores techniques in machine learning.",
"interest_category": "Offline-First / Local AI"
},
{
"title": "TabTune: A Unified Library for Inference and Fine-Tuning Tabular Foundation Models",
"summary": " TabTune provides consistent access to seven state-of-the-art models supporting multiple adaptation strategies . The framework automates model-aware preprocessing, manages architectural heterogeneity internally, and integrates evaluation modules for performance, calibration, and fairness . The library is open source and available at https://github.com/Lexsi-Labs/TabTune .",
"link": "http://arxiv.org/abs/2511.02802v1",
"pdf_link": "https://arxiv.org/pdf/2511.02802.pdf",
"arxiv_id": "2511.02802",
"category": "cs.LG",
"published": "2025-11-04",
"relevance_score": 4,
"matched_keywords": [
"efficient",
"inference"
],
"difficulty": "🟢 Applied",
"layman": "This research presents techniques for machine learning.",
"interest_category": "Efficient ML / Edge AI"
},
{
"title": "Assessing win strength in MLB win prediction models",
"summary": " In Major League Baseball, strategy and planning are major factors in determining the outcome of a game . Previous studies have aided this by building machine learning models for predicting the winning team of any given game . We extend this work by training a comprehensive set of models using a common dataset .",
"link": "http://arxiv.org/abs/2511.02815v1",
"pdf_link": "https://arxiv.org/pdf/2511.02815.pdf",
"arxiv_id": "2511.02815",
"category": "cs.LG",
"published": "2025-11-04",
"relevance_score": 1,
"matched_keywords": [
"differential"
],
"difficulty": "🟢 Applied",
"layman": "This research explores techniques in machine learning.",
"interest_category": "Privacy-Preserving ML"
},
{
"title": "Neurosymbolic Deep Learning Semantics",
"summary": " Artificial Intelligence (AI) is a powerful new language of science as evidenced by recent Nobel Prizes in chemistry and physics . Yet, this new language lacks semantics, which makes AI's scientific discoveries unsatisfactory at best . In this paper, we argue that logic offers an adequate framework to offer a much needed semantics for deep learning .",
"link": "http://arxiv.org/abs/2511.02825v1",
"pdf_link": "https://arxiv.org/pdf/2511.02825.pdf",
"arxiv_id": "2511.02825",
"category": "cs.AI",
"published": "2025-11-04",
"relevance_score": 2,
"matched_keywords": [
"art",
"design"
],
"difficulty": "🟢 Applied",
"layman": "This research explores techniques in edge computing.",
"interest_category": "Creative AI / Emotion"
},
{
"title": "3D Point Cloud Object Detection on Edge Devices for Split Computing",
"summary": " Split Computing aims to lessen the computational burden on edge devices, thereby reducing processing time and power consumption . Split Computing reduces the inference time by 70.8% and the edge device execution time by 90.0% .",
"link": "http://arxiv.org/abs/2511.02293v1",
"pdf_link": "https://arxiv.org/pdf/2511.02293.pdf",
"arxiv_id": "2511.02293",
"category": "cs.DC",
"published": "2025-11-04",
"relevance_score": 6,
"matched_keywords": [
"edge",
"device"
],
"difficulty": "🟢 Applied",
"layman": "This research automatically finding edge computing.",
"interest_category": "Lightweight Systems"
},
{
"title": "From Solo to Symphony: Orchestrating Multi-Agent Collaboration with Single-Agent Demos",
"summary": " Solo-to-Collaborative RL (SoCo) is a framework that transfers solo knowledge into cooperative learning . SoCo pretrains a shared solo policy from solo demonstrations, then adapts it for cooperation during multi-agent training . It significantly boosts training efficiency and performance of backbone algorithms .",
"link": "http://arxiv.org/abs/2511.02762v1",
"pdf_link": "https://arxiv.org/pdf/2511.02762.pdf",
"arxiv_id": "2511.02762",
"category": "cs.LG",
"published": "2025-11-04",
"relevance_score": 2,
"matched_keywords": [
"offline",
"edge"
],
"difficulty": "🟢 Applied",
"layman": "This research makes more efficient edge computing.",
"interest_category": "Offline-First / Local AI"
},
{
"title": "Does Interpretability of Knowledge Tracing Models Support Teacher Decision Making?",
"summary": " Knowledge tracing (KT) models are a crucial basis for pedagogical decision-making . No study to date has investigated whether the interpretability of KT models actually helps human teachers to make teaching decisions .",
"link": "http://arxiv.org/abs/2511.02718v1",
"pdf_link": "https://arxiv.org/pdf/2511.02718.pdf",
"arxiv_id": "2511.02718",
"category": "cs.LG",
"published": "2025-11-04",
"relevance_score": 3,
"matched_keywords": [
"edge"
],
"difficulty": "🟢 Applied",
"layman": "This research running AI locally on devices for edge computing.",
"interest_category": "Efficient ML / Edge AI"
},
{
"title": "TWIST2: Scalable, Portable, and Holistic Humanoid Data Collection System",
"summary": " TWIST2 is a portable, mocap-free humanoid teleoperation and data collection system . System leverages PICO4U VR for obtaining real-time whole-body human motions . System is fully reproducible and open-sourced at https://yanjieze.com/TWIST2 .",
"link": "http://arxiv.org/abs/2511.02832v1",
"pdf_link": "https://arxiv.org/pdf/2511.02832.pdf",
"arxiv_id": "2511.02832",
"category": "cs.RO",
"published": "2025-11-04",
"relevance_score": 0,
"matched_keywords": [],
"difficulty": "🟢 Applied",
"layman": "This research explores techniques in language AI.",
"interest_category": "Privacy-Preserving ML"
},
{
"title": "Perceived Femininity in Singing Voice: Analysis and Prediction",
"summary": " This paper focuses on the often-overlooked aspect of perceived voice femininity in singing voices . The analysis of gender bias in music content could benefit from such study .",
"link": "http://arxiv.org/abs/2511.02726v1",
"pdf_link": "https://arxiv.org/pdf/2511.02726.pdf",
"arxiv_id": "2511.02726",
"category": "cs.SD",
"published": "2025-11-04",
"relevance_score": 4,
"matched_keywords": [
"music",
"art",
"design"
],
"difficulty": "🟢 Applied",
"layman": "This research explores techniques in speech processing.",
"interest_category": "Creative AI / Emotion"
},
{
"title": "FREESH: Fair, Resource- and Energy-Efficient Scheduling for LLM Serving on Heterogeneous GPUs",
"summary": " The ever-increasing computation and energy demand for LLM and AI agents call for holistic and efficient optimization of LLM serving systems . FREESH identifies optimal configurations of balanced load serving by matching distinct GPU instance's power-throughput characteristics with predictable LLM query length and workloads . During the 1-hour serving on production workloads, FREESH reduces energy by 28.6% and emissions by 45.45% .",
"link": "http://arxiv.org/abs/2511.00807v1",
"pdf_link": "https://arxiv.org/pdf/2511.00807.pdf",
"arxiv_id": "2511.00807",
"category": "cs.DC",
"published": "2025-11-02",
"relevance_score": 4,
"matched_keywords": [
"iot",
"resource"
],
"difficulty": "🟢 Applied",
"layman": "This research optimizes language AI.",
"interest_category": "Lightweight Systems"
},
{
"title": "A Non-Adversarial Approach to Idempotent Generative Modelling",
"summary": " Idempotent Generative Networks (IGNs) are deep generative models that also function as local data manifold projectors, mapping arbitrary inputs back onto the manifold . IGNs suffer from mode collapse, mode dropping, and training instability due to their objectives .",
"link": "http://arxiv.org/abs/2511.02614v1",
"pdf_link": "https://arxiv.org/pdf/2511.02614.pdf",
"arxiv_id": "2511.02614",
"category": "cs.LG",
"published": "2025-11-04",
"relevance_score": 1,
"matched_keywords": [
"local"
],
"difficulty": "🟢 Applied",
"layman": "This research creating new content with machine learning.",
"interest_category": "Offline-First / Local AI"
}
];
// ============================================
// STATE MANAGEMENT
// ============================================
let likes = JSON.parse(localStorage.getItem('tiktok_likes') || '{}');
let currentPaperIndex = 0;
// ============================================
// RENDER FEED
// ============================================
function renderFeed() {
const container = document.getElementById('feed-container');
papers.forEach((paper, index) => {
const card = document.createElement('div');
card.className = 'paper-card';
card.dataset.index = index;
card.innerHTML = `
<div class="interest-badge">${paper.interest_category}</div>
<div class="difficulty-badge">${paper.difficulty}</div>
<h1 class="paper-title">${paper.title}</h1>
<div class="layman-box">💡 ${paper.layman}</div>
<div class="summary">${paper.summary}</div>
<div class="paper-meta">
<span class="category-tag">${paper.category}</span>
<span class="date">${paper.published}</span>
</div>
<div class="links">
<a href="${paper.link}" target="_blank">Abstract ↗</a>
<a href="${paper.pdf_link}" target="_blank">PDF ↗</a>
</div>
`;
container.appendChild(card);
});
}
// ============================================
// LIKE SYSTEM
// ============================================
function getCurrentPaper() {
const container = document.getElementById('feed-container');
const scrollPos = container.scrollTop;
const windowHeight = window.innerHeight;
// Find which paper is currently in view
const cards = document.querySelectorAll('.paper-card');
for (let i = 0; i < cards.length; i++) {
const rect = cards[i].getBoundingClientRect();
if (rect.top >= -windowHeight/2 && rect.top < windowHeight/2) {
return i;
}
}
return 0;
}
function toggleLike() {
const paperIndex = getCurrentPaper();
const paper = papers[paperIndex];
const arxivId = paper.arxiv_id;
const heartIcon = document.getElementById('heartIcon');
const likeButton = document.getElementById('likeButton');
if (likes[arxivId]) {
// Unlike
delete likes[arxivId];
heartIcon.textContent = '♡';
likeButton.classList.remove('liked');
} else {
// Like
likes[arxivId] = {
arxiv_id: arxivId,
title: paper.title,
abstract_url: paper.link,
category: paper.category,
interest_category: paper.interest_category,
liked_date: new Date().toISOString(),
difficulty: paper.difficulty
};
heartIcon.textContent = '♥';
likeButton.classList.add('liked');
}
// Save to localStorage
localStorage.setItem('tiktok_likes', JSON.stringify(likes));
// Update counter and export button
updateCounter();
updateExportButton();
}
function updateLikeButton() {
const paperIndex = getCurrentPaper();
const paper = papers[paperIndex];
const heartIcon = document.getElementById('heartIcon');
const likeButton = document.getElementById('likeButton');
if (likes[paper.arxiv_id]) {
heartIcon.textContent = '♥';
likeButton.classList.add('liked');
} else {
heartIcon.textContent = '♡';
likeButton.classList.remove('liked');
}
}
function updateCounter() {
const count = Object.keys(likes).length;
document.getElementById('likeCount').textContent = count;
}
function updateExportButton() {
const exportButton = document.getElementById('exportButton');
if (Object.keys(likes).length > 0) {
exportButton.classList.add('active');
} else {
exportButton.classList.remove('active');
}
}
// ============================================
// EXPORT LIKES
// ============================================
function exportLikes() {
const likedPapers = Object.values(likes);
// Calculate category preferences
const preferences = {};
likedPapers.forEach(paper => {
const cat = paper.interest_category;
preferences[cat] = (preferences[cat] || 0) + 1;
});
const exportData = {
liked_papers: likedPapers,
preferences: preferences,
export_date: new Date().toISOString(),
total_likes: likedPapers.length
};
const blob = new Blob([JSON.stringify(exportData, null, 2)], {
type: 'application/json'
});
const url = URL.createObjectURL(blob);
const a = document.createElement('a');
a.href = url;
a.download = `arxiv_likes_${new Date().toISOString().split('T')[0]}.json`;
document.body.appendChild(a);
a.click();
document.body.removeChild(a);
URL.revokeObjectURL(url);
}
// ============================================
// EVENT LISTENERS
// ============================================
document.getElementById('likeButton').addEventListener('click', toggleLike);
document.getElementById('exportButton').addEventListener('click', exportLikes);
// Update like button when scrolling
document.getElementById('feed-container').addEventListener('scroll', () => {
updateLikeButton();
// Hide scroll indicator after first scroll
const scrollIndicator = document.getElementById('scrollIndicator');
if (document.getElementById('feed-container').scrollTop > 50) {
scrollIndicator.classList.add('hide-indicator');
}
});
// ============================================
// INITIALIZATION
// ============================================
renderFeed();
updateLikeButton();
updateCounter();
updateExportButton();
</script>
</body>
</html>
Reference in New Issue View Git Blame Copy Permalink