Friday, May 1, 2026
Latest

Archive — Page 2

67 articles total
OpenAI Launches GPT-Rosalind for Drug Discovery and Genomics

OpenAI Launches GPT-Rosalind for Drug Discovery and Genomics

Frontier reasoning model targets fragmented workflows in life sciences research with limited access release.

Two studies reveal mechanisms behind transformer grokking delays

Two studies reveal mechanisms behind transformer grokking delays

New research identifies why neural networks suddenly generalize long after memorizing training data.

Frontier AI Agents Fail One in Three Production Attempts

Frontier AI Agents Fail One in Three Production Attempts

Stanford HAI and enterprise research reveal systemic reliability gaps in deployed agentic systems.

Linear Programming Removes Subjectivity From Multi-Criteria Decisions

Linear Programming Removes Subjectivity From Multi-Criteria Decisions

New method combines cardinal and ordinal data to rank alternatives without human bias distorting results.

WSU Researchers Deploy AI Spectral Imaging for Plastic Sorting

WSU Researchers Deploy AI Spectral Imaging for Plastic Sorting

Spectral imaging combined with machine learning improves automated identification of recyclable plastic types.

LLMs Learn Backwards: Scaling Hypothesis Faces Hard Limits

LLMs Learn Backwards: Scaling Hypothesis Faces Hard Limits

New analysis suggests large language models acquire intelligence in reverse order, challenging unlimited scaling assumptions.

AI Discourse Splits Over Capability Claims and Economic Impact

AI Discourse Splits Over Capability Claims and Economic Impact

Leaders and practitioners report diverging assessments of current AI reliability and societal effects.

Prediction Arena Lets AI Models Trade Real Capital on Live Markets

Prediction Arena Lets AI Models Trade Real Capital on Live Markets

Researchers introduce first benchmark that tests AI decision-making against actual prediction market exchanges.

GLM 5.1 Benchmarks Show Performance Across Major Model Tasks

GLM 5.1 Benchmarks Show Performance Across Major Model Tasks

Zhipu AI's latest model demonstrates measurable capability gains in standardized evaluation protocols.

Hybrid Attention Architecture Achieves 50x Faster Code Model Inference

Hybrid Attention Architecture Achieves 50x Faster Code Model Inference

A researcher modified transformer attention mechanisms to dramatically speed up inference on small language models.

KDD and NeurIPS Review Cycles Collide in Early 2026

KDD and NeurIPS Review Cycles Collide in Early 2026

Multiple top-tier ML conferences releasing decisions simultaneously creates unusual timing pressures for researchers.

Anthropic Discovers 171 Emotion Vectors Steering Claude Behavior

Anthropic Discovers 171 Emotion Vectors Steering Claude Behavior

Researchers identify functional emotional patterns in Claude's neural activations that influence decision-making.