AllLLMRoboticsStartupsResearchEthicsBusinessHardwareAutonomousMedicineCreative
Topic

#LLM

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

4859 articles-78% vs prev. week296 this week
Subscribe
RELATED:#Business#Hardware#Research#Creative#Startups#Ethics#Autonomous#Robotics#Medicine
Researchers Prove Why Adversarial Training Makes Saliency Maps Sparse
LLM·arXiv cs.LG

Researchers Prove Why Adversarial Training Makes Saliency Maps Sparse

Oct 9, 2026·2 min
Subliminal learning passed a backdoor through numbers: 23.5% of replies in French
LLM

Subliminal learning passed a backdoor through numbers: 23.5% of replies in French

Oct 9, 2026·2 min
PXtal Beats Baseline Models on Six Benchmarks, Four of Them Zero-Shot
LLM

PXtal Beats Baseline Models on Six Benchmarks, Four of Them Zero-Shot

Oct 9, 2026·2 min
Nullify Erases Knowledge From LLMs Without Retraining, Using Null-Space Vectors
LLM

Nullify Erases Knowledge From LLMs Without Retraining, Using Null-Space Vectors

Oct 9, 2026·2 min
CARing teaches LLMs to predict diagnoses: R@30 hits 46.52% on MIMIC-IV
LLM·arXiv cs.LG

CARing teaches LLMs to predict diagnoses: R@30 hits 46.52% on MIMIC-IV

CARing's authors encode diagnoses as Semantic IDs and reward the model for covering distinct diagnoses. In reasoning mode, R@30 is 46.04% on MIMIC-III and 46.52% on…

Oct 9, 2026·2 min
Visible CoT Does Not Universally Improve SQL and pandas Code Generation, New Study Finds
LLM·arXiv cs.LG

Visible CoT Does Not Universally Improve SQL and pandas Code Generation, New Study Finds

An arXiv preprint tested the "think out loud" advice on SQL and pandas with code execution. The authors found no general CoT gain and no benefit from "thinking in…

Oct 9, 2026·2 min
D-SLR Replaces Truncated SVD With Three Decompositions and No Tuning
LLM·arXiv cs.LG

D-SLR Replaces Truncated SVD With Three Decompositions and No Tuning

D-SLR, a new matrix compression method, stores some rows verbatim and approximates the rest with a low-rank fit. Three SVDs, zero hyperparameters, and a guarantee…

Oct 9, 2026·2 min
Phase-HDC trains a model with no optimizer history and stores 16–23x less than Adam
LLM·arXiv cs.LG

Phase-HDC trains a model with no optimizer history and stores 16–23x less than Adam

Phase-HDC replaces Adam's moments with a threshold on the current gradient and stores 16–23x less than float32 Adam. The cost: about five points of accuracy on…

Oct 9, 2026·3 min
Top themes
01IKNO for Navier–Stokes Stays Accurate Under Rotation, With 5.6x Fewer Parameters Than TFNOct 9, 2026·2 min02LoopOPD Teaches Looped Language Models to Improve Through Their Own LoopsOct 9, 2026·2 min03MoE Router Telemetry Leaks Fine-Tuning Data: Attack Gains 2.7–9.4 pp in AccuracyOct 9, 2026·3 min0480 Synthetic Dialogues Are Enough to Train Activation Probes for Gemma-3-27B, but Not EverywhereOct 9, 2026·2 min05Temporal Transformer for the CAN Bus Catches Anomalies Without Sharing Raw DataOct 9, 2026·2 min
Freezing the Decoder in KV-Cache Compression Saves 3x Memory but Doesn't Improve Accuracy
LLM·arXiv cs.LG

Freezing the Decoder in KV-Cache Compression Saves 3x Memory but Doesn't Improve Accuracy

Oct 9, 2026·2 min
SFT-as-Context Restores General Skills in Fine-Tuned LLMs, Within 2.2 Points
LLM·arXiv cs.CL

SFT-as-Context Restores General Skills in Fine-Tuned LLMs, Within 2.2 Points

Oct 9, 2026·3 min
Lapras replaces CoT with latent reasoning: up to +10.79% F1 with 23.9x fewer tokens
LLM·arXiv cs.CL

Lapras replaces CoT with latent reasoning: up to +10.79% F1 with 23.9x fewer tokens

Oct 9, 2026·2 min
Doctors Don't Ask LLMs About Diagnoses: 127 833 Queries vs. 58 Benchmarks
LLM·arXiv cs.CL

Doctors Don't Ask LLMs About Diagnoses: 127 833 Queries vs. 58 Benchmarks

Oct 9, 2026·2 min
FedAlphaEdit Merges Knowledge Edits from Different Organizations into One LLM Without Sharing Data
LLM·arXiv cs.CL

FedAlphaEdit Merges Knowledge Edits from Different Organizations into One LLM Without Sharing Data

Researchers show that naively combining AlphaEdit and CollabEdit breaks down. Their FedAlphaEdit merges edits in a shared null space and comes close to centralized…

Oct 9, 2026·2 min
Three LLMs Follow Prompts for Stutters and Interruptions Poorly; Rules Do Better
LLM·arXiv cs.CL

Three LLMs Follow Prompts for Stutters and Interruptions Poorly; Rules Do Better

The tau-Voice authors tested how three LLMs imitate disfluencies, interruptions and backchanneling in user simulators. A deterministic algorithm proved more…

Oct 9, 2026·2 min
Sociolinguistics Instead of LLM Judges: Testing User-Simulator Personas on 5 Agents
LLM·arXiv cs.CL

Sociolinguistics Instead of LLM Judges: Testing User-Simulator Personas on 5 Agents

A pilot arXiv study defines simulator personas through measurable style parameters and checks them with stylometry on five customer-service agents.

Oct 9, 2026·2 min
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·arXiv cs.CL

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·arXiv cs.CL

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·arXiv cs.CL

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

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

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

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

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

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

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

Oct 9, 2026·2 min
Text autoencoder compresses web text to 2.24 bits per byte without losing meaning
LLM·arXiv cs.CL

Text autoencoder compresses web text to 2.24 bits per byte without losing meaning

A new arXiv paper trains a neural network to compress text into a discrete latent: 2.24 bits per byte on web data, on par with lossless algorithms, with good…

Oct 9, 2026·2 min
Cognitive Thermometers: Neural Networks Proposed as a Way to Measure the Complexity of Meaning
LLM·arXiv cs.CL

Cognitive Thermometers: Neural Networks Proposed as a Way to Measure the Complexity of Meaning

A survey on arXiv proposes using machine learning models as "cognitive thermometers" to measure which meanings languages choose most often, and why.

Oct 9, 2026·2 min
Related topics
Business 1160Research 516Robotics 178Hardware 106Autonomous 33Startups 9Medicine 8Ethics 8