AllLLMRoboticsStartupsResearchEthicsBusinessHardwareAutonomousMedicineCreative
Topic

#Research

Curated coverage on «Research»: releases, research, benchmarks, open-source.

1209 articles-55% vs prev. week146 this week
Subscribe
RELATED:#LLM#Business#Hardware#Creative#Startups#Ethics#Autonomous#Robotics#Medicine
Neural Networks Are More Modular Than Their Neurons Suggest: Three Levels of Testing
Research·arXiv cs.AI

Neural Networks Are More Modular Than Their Neurons Suggest: Three Levels of Testing

Oct 12, 2026·2 min
New Method Aligns Encoders With 100–500 Pairs Instead of Millions
Research

New Method Aligns Encoders With 100–500 Pairs Instead of Millions

Oct 12, 2026·2 min
LLM Pipeline Turns Datasheets into Knowledge Graphs with 98.8% Accuracy
Research

LLM Pipeline Turns Datasheets into Knowledge Graphs with 98.8% Accuracy

Oct 12, 2026·2 min
EvoSyn: Multi-Agent Data Synthesis Lifts Task Success by 13.3 Points
Research

EvoSyn: Multi-Agent Data Synthesis Lifts Task Success by 13.3 Points

Oct 12, 2026·2 min
MCTS Loses to Random Restarts by 2.88% in DAG Scheduling at Equal Budget
Research·arXiv cs.AI

MCTS Loses to Random Restarts by 2.88% in DAG Scheduling at Equal Budget

The authors gave MCTS and a simple list scheduler with restarts 2048 schedules each on DAG tasks. The search tree came out 2.88% worse than random restarts.

Oct 12, 2026·2 min
arXiv paper finds five flaws in XAI faithfulness metrics
Research·arXiv cs.AI

arXiv paper finds five flaws in XAI faithfulness metrics

An arXiv position paper says a decade of XAI has not raised trust in models, and faithfulness metrics have five systemic flaws. The authors propose four research…

Oct 12, 2026·2 min
Zebra3D speeds up 3D reconstruction by up to 7.1× on 1024 frames without training from scratch
Research·arXiv cs.CV

Zebra3D speeds up 3D reconstruction by up to 7.1× on 1024 frames without training from scratch

Zebra3D's hybrid attention speeds up off-the-shelf VGGT, HunyuanWorld-Mirror and π³ by up to 7.1× on 1024 views. At 94% sparsity, VGGT accuracy drops by only 1.1 pp.

Oct 12, 2026·2 min
STAPLE adds billions of parameters to OLMo-2 and decodes twice as fast as MoE
Research·arXiv cs.LG

STAPLE adds billions of parameters to OLMo-2 and decodes twice as fast as MoE

STAPLE embeds static sparse memory into a pretrained LLM. On OLMo-2 1B it gains +4.27 points on HumanEval, while training and generation slow by no more than 3%.

Oct 12, 2026·2 min
Top themes
01MisVisAgentBench: Misleading Charts Cut Web Agent Success by 34 Percentage PointsOct 12, 2026·2 min02Conversational Model Benchmark: GPT Realtime and Gemini Respond at Almost Every PauseOct 12, 2026·2 min03Linear Head on Qwen3-VL Beats GPT-4o at Scoring Answers Without Human LabelsOct 12, 2026·2 min04GEPA Lifts a Deep-Research Agent's Reward from 0.702 to 0.781 on Unseen TopicsOct 12, 2026·2 min05ELR teaches one model to loop: Koi-4B scores 18 and 29Oct 12, 2026·2 min
DCOCO speeds up LLM inference by up to 6.6x and beats rivals by 6–23%
Research·arXiv cs.CL

DCOCO speeds up LLM inference by up to 6.6x and beats rivals by 6–23%

Oct 12, 2026·2 min
EvoBelief teaches LLM agents to test hypotheses about the world: 83.3% vs 0% for baselines
Research·arXiv cs.AI

EvoBelief teaches LLM agents to test hypotheses about the world: 83.3% vs 0% for baselines

Oct 12, 2026·2 min
MagicBeanStalk lifts gpt-oss-120b from 61.2 to 69.8 points on math
Research·arXiv cs.AI

MagicBeanStalk lifts gpt-oss-120b from 61.2 to 69.8 points on math

Oct 12, 2026·2 min
SpireBench Puts 20 LLMs Through a Full Run of Slay the Spire: Experts Still Ahead
Research·arXiv cs.AI

SpireBench Puts 20 LLMs Through a Full Run of Slay the Spire: Experts Still Ahead

Oct 12, 2026·2 min
Dream2Discern Predicts Multi-Agent System Failures With Over 90% Accuracy
Research·Hugging Face Daily Papers

Dream2Discern Predicts Multi-Agent System Failures With Over 90% Accuracy

The Dream2Discern auditor trains on "hard" latent variants of already-recorded trajectories and reaches over 90% accuracy on AFTraj-2K and StepShield without new…

Oct 12, 2026·2 min
TabWM learns to predict database edits: F1 rises from 0.016 to 0.685
Research·Hugging Face Daily Papers

TabWM learns to predict database edits: F1 rises from 0.016 to 0.685

From 64 execution examples, TabWM predicts how an action will change the relational database of an unfamiliar system: full edit F1 of 0.685 versus 0.016 with no…

Oct 12, 2026·3 min
Why Repeated Fine-Tuning Hits a Ceiling: The LANN-4 Study
Research·Habr AI

Why Repeated Fine-Tuning Hits a Ceiling: The LANN-4 Study

The LANN-4 author ran experiments with repeated fine-tuning of a small model. The first round produced a jump from 53% to 59% correct answers on 200 tasks — the next three rounds added nothing. The op

Oct 11, 2026·3 min
DeepSeek-OCR Compresses Text Context 10x at 97% Accuracy
Research·YouTube: Two Minute Papers

DeepSeek-OCR Compresses Text Context 10x at 97% Accuracy

Oct 11, 2026·2 min
Ten Minutes With AI Reduce Persistence on Hard Problems, Berkeley Study Finds
Research·Hacker News AI

Ten Minutes With AI Reduce Persistence on Hard Problems, Berkeley Study Finds

Oct 11, 2026·2 min
Terence Tao Gives Caltech Lecture "Math 2.0" on Machine-Assisted Proofs
Research·Jiqizhixin (机器之心)

Terence Tao Gives Caltech Lecture "Math 2.0" on Machine-Assisted Proofs

Oct 11, 2026·2 min
arXiv caps authors at two papers a month as AI submissions surge
Research·The Decoder

arXiv caps authors at two papers a month as AI submissions surge

Oct 11, 2026·2 min
Epoch AI: Recent Models Struggled to Match a Human-Invented Algorithmic Idea
Research·Hacker News AI

Epoch AI: Recent Models Struggled to Match a Human-Invented Algorithmic Idea

Oct 11, 2026·2 min
Odyssey-3: how a video model teaches robots physics in tens of hours
Research

Odyssey-3: how a video model teaches robots physics in tens of hours

Oct 11, 2026·3 min
Hacker-fixer loop closes holes in the verifiers of Terminal Bench, KernelBench and SETA
Research

Hacker-fixer loop closes holes in the verifiers of Terminal Bench, KernelBench and SETA

Oct 11, 2026·2 min
ReviewScope hunts for anomalies in reviews using embeddings and graphs across 21,000 texts
Research

ReviewScope hunts for anomalies in reviews using embeddings and graphs across 21,000 texts

Oct 10, 2026·2 min
Can AI Spot a Network "Black Hole" Before It Fails? A Look at Silent Packet Loss
Research·TechXplore AI

Can AI Spot a Network "Black Hole" Before It Fails? A Look at Silent Packet Loss

A router looks healthy but silently drops some packets. Why these failures go unnoticed, and what AI could contribute to finding them.

Oct 10, 2026·2 min
Hebbian Learning Without Backprop Collapsed Into k-means and Lost on Every Front
Research·Habr AI

Hebbian Learning Without Backprop Collapsed Into k-means and Lost on Every Front

A Habr author trained a hidden layer using only local Hebbian rules. The rule collapsed into k-means and beat backprop on accuracy, on few-label performance and on…

Oct 10, 2026·2 min
Related topics
LLM 1628Business 1117Robotics 174Hardware 100Autonomous 17Startups 9Medicine 8Ethics 5