AllLLMRoboticsStartupsResearchEthicsBusinessHardwareAutonomousMedicineCreative
Topic

#LLM

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

4839 articles-77% vs prev. week324 this week
Subscribe
RELATED:#Business#Hardware#Research#Creative#Startups#Ethics#Autonomous#Robotics#Medicine
AI4Fire: SQL Tool Lifts LLM Accuracy on Wildfire Databases From 16% to 88%
LLM·arXiv cs.CL

AI4Fire: SQL Tool Lifts LLM Accuracy on Wildfire Databases From 16% to 88%

Oct 9, 2026·2 min
LLMs Judge Translation Across 70 Language Pairs: They Beat Humans Only in Places
LLM

LLMs Judge Translation Across 70 Language Pairs: They Beat Humans Only in Places

Oct 9, 2026·2 min
STI-OPD: A Probabilistic Teacher Intervention for Multi-Step Agentic Tasks
LLM

STI-OPD: A Probabilistic Teacher Intervention for Multi-Step Agentic Tasks

Oct 9, 2026·2 min
MASA Reframes LLM Sparse Attention as Matrix Approximation
LLM

MASA Reframes LLM Sparse Attention as Matrix Approximation

Oct 9, 2026·2 min
Routed Sparse Autoencoders Separate Words From Intonation in Speech on WavLM and SPEAR
LLM·arXiv cs.CL

Routed Sparse Autoencoders Separate Words From Intonation in Speech on WavLM and SPEAR

A new TopK autoencoder method sends speech content and voice with emotion down separate routes. Tested on two encoders, two corpora and independent probes.

Oct 9, 2026·2 min
Galahad-kv stores KV cache on NVMe and loads a block 2.8–4.3x faster than recomputing it
LLM·arXiv cs.CL

Galahad-kv stores KV cache on NVMe and loads a block 2.8–4.3x faster than recomputing it

The galahad-kv package stores the KV state of 50M tokens on encrypted NVMe. On an H100, loading a block is 2.8–4.3x faster than recomputing it and uses 8.8–12.3x…

Oct 9, 2026·2 min
Fine-Tuned 0.8B-Parameter Model Beats GPT-5.4 at Grammar Annotation
LLM·arXiv cs.CL

Fine-Tuned 0.8B-Parameter Model Beats GPT-5.4 at Grammar Annotation

A 0.8B Qwen3.5 fine-tuned on teacher-model labels beat GPT-5.4 and GPT-5.6 Sol on precision and recall, and cut inference costs roughly 16x.

Oct 9, 2026·2 min
Utterance Normalization Cuts the Cost of Analyzing Millions of Calls: A Pipeline Built Entirely on Small Models
LLM·arXiv cs.CL

Utterance Normalization Cuts the Cost of Analyzing Millions of Calls: A Pipeline Built Entirely on Small Models

An arXiv paper turns dialogues into short, attributed statements with tags. A shared preparation layer lets teams run call analytics on small models.

Oct 9, 2026·2 min
Top themes
01Text autoencoder compresses web text to 2.24 bits per byte without losing meaningOct 9, 2026·2 min02Cognitive Thermometers: Neural Networks Proposed as a Way to Measure the Complexity of MeaningOct 9, 2026·2 min03LLMs Draft Road-Work Plans for WisDOT: 14B Barely Beats 7BOct 9, 2026·2 min04Wieszcz-XIX: A 3.1-Billion-Word Polish Corpus From Before 1918 Trains Models From ScratchOct 9, 2026·2 min05Diffu-LoRA Lets the Adapter Decide Which Stable Diffusion Layers Need RankOct 9, 2026·2 min
GenUI-Harness Lifts a 4B Model From 9.33% to 58% Pass@3, Beating Claude Opus 5
LLM·arXiv cs.AI

GenUI-Harness Lifts a 4B Model From 9.33% to 58% Pass@3, Beating Claude Opus 5

Oct 9, 2026·2 min
OpenProblemBench: GPT-6-Astra Solves 14% of Open Math and Physics Problems
LLM·arXiv cs.AI

OpenProblemBench: GPT-6-Astra Solves 14% of Open Math and Physics Problems

Oct 9, 2026·2 min
AgentHorizon: Best Judge GPT-5.5 Scores Only 80.9% on Long Agent Tasks
LLM·arXiv cs.AI

AgentHorizon: Best Judge GPT-5.5 Scores Only 80.9% on Long Agent Tasks

Oct 9, 2026·2 min
Two Distillation Schemes Make AI Students Reveal Their Teachers' Hidden Behavior
LLM·arXiv cs.AI

Two Distillation Schemes Make AI Students Reveal Their Teachers' Hidden Behavior

Oct 9, 2026·2 min
Researchers prove that bloating AGENTS.md is harmful: the file has a size limit
LLM·arXiv cs.AI

Researchers prove that bloating AGENTS.md is harmful: the file has a size limit

Preprint arXiv 2610.11007 models an agent's context file as a capacitated assortment problem and proves an upper bound on its optimal size.

Oct 9, 2026·2 min
GUISE Shows Narrative Wrappers Break Qwen3-1.7B in 89–96% of Cases
LLM·arXiv cs.AI

GUISE Shows Narrative Wrappers Break Qwen3-1.7B in 89–96% of Cases

Researchers built the GUISE benchmark: a role-play wrapper bypasses Qwen3-1.7B's safeguards 89.4% of the time in English and 95.7% in Classical Chinese. They also…

Oct 9, 2026·2 min
An LLM Teaches a Planner to Get by With Polynomial Memory and Solves 1709 of 1890 Tasks
LLM·arXiv cs.AI

An LLM Teaches a Planner to Get by With Polynomial Memory and Solves 1709 of 1890 Tasks

Researchers replaced storage of millions of states with a learned policy with registers. Search runs in polynomial memory and solves 1709 of 1890 IPC 2023 test tasks.

Oct 9, 2026·2 min
StoreBench: LLM Agents Run an Online Store and Lose to a Script — 49% vs. 97%
LLM·arXiv cs.AI

StoreBench: LLM Agents Run an Online Store and Lose to a Script — 49% vs. 97%

Oct 9, 2026·3 min
LLM Agents Don't Learn Community Norms From Interaction, Study Finds
LLM·arXiv cs.AI

LLM Agents Don't Learn Community Norms From Interaction, Study Finds

Oct 9, 2026·2 min
Qwen3.6-27B and Qwen3-4B Agents Blow Their Time Budgets Without Harness Hints
LLM·arXiv cs.AI

Qwen3.6-27B and Qwen3-4B Agents Blow Their Time Budgets Without Harness Hints

Oct 9, 2026·2 min
Plan-and-Patch: Diffusion LLM Repairs Agent Plans Twice as Often as Qwen3-8B
LLM·arXiv cs.AI

Plan-and-Patch: Diffusion LLM Repairs Agent Plans Twice as Often as Qwen3-8B

Oct 9, 2026·2 min
arXiv authors call for replacing a single annotation "truth" with a space of interpretations
LLM·arXiv cs.AI

arXiv authors call for replacing a single annotation "truth" with a space of interpretations

Oct 9, 2026·2 min
TGIT Translates Short Drone Commands and Lifts Navigation Success from 11% to 33%
LLM

TGIT Translates Short Drone Commands and Lifts Navigation Success from 11% to 33%

Oct 9, 2026·2 min
Agent strikes out its own old tool outputs and cuts input tokens by half
LLM

Agent strikes out its own old tool outputs and cuts input tokens by half

Oct 9, 2026·2 min
Credit AI Agent Can Only Change Its Harness: Gate Admitted 144 of 7,449 Edits
LLM

Credit AI Agent Can Only Change Its Harness: Gate Admitted 144 of 7,449 Edits

Oct 9, 2026·3 min
arXiv Authors Bound Agent Self-Improvement by Verification Class
LLM·arXiv cs.AI

arXiv Authors Bound Agent Self-Improvement by Verification Class

A theoretical preprint separates three ways to improve an agent: search longer, get more support, or change verification. The complexity classes are sandwiched…

Oct 9, 2026·2 min
Western open-weight models pitch businesses an alternative to Chinese and closed AI
LLM·AI Business

Western open-weight models pitch businesses an alternative to Chinese and closed AI

Western open-weight models are pitched to companies as an alternative to Chinese and proprietary AI. They offer more control, but the customer takes on…

Oct 9, 2026·2 min
Related topics
Business 1160Research 499Robotics 178Hardware 101Autonomous 17Startups 9Medicine 8Ethics 5