Ground Truth.
AI, checked against the source.
← 2026-08-142026-08-15later →

MIT found brain-like modules inside six large language models

2026-08-15

MIT researchers localized the neurons behind 46 reasoning tasks in six large language models and found that tasks sharing a brain network in humans share neurons in the models, with 4.3 times more overlap within a cognitive domain than across domains.

interpretability · mechanistic-interpretability · neuroscience · research · llm

A 9B model writes agent upgrades as good as Claude Opus 4.6

2026-08-15

A 17-author study separates the ability to improve an AI agent's scaffolding from the ability to benefit from the improvement, and finds that a 9-billion-parameter model produces upgrades yielding gains comparable to Claude Opus 4.6.

agents · agent-harness · self-improvement · research · evaluation

Agent skill libraries now need a librarian, not a folder

2026-08-15

SkillsVote profiles a million-scale corpus of open-source agent skills and admits new ones only through evidence-gated updates after execution, arguing that indiscriminate additions to a skill library pollute every future run's context.

agents · agent-skills · context-engineering · research · coding-agents

Compressed memory stretched a 7,000-token model to 1.75 million

2026-08-15

A framework that chunks long documents, compresses each chunk into memory blocks and gates which blocks the model reads extrapolated from 7,000 tokens of training context to 1.75 million at inference, with half the peak GPU memory of a leading baseline.

long-context · memory · reinforcement-learning · efficiency · research

EXO keeps an agent's memory outside the code the agent rewrites

2026-08-15

The EXO agent runtime splits a self-modifying agent into a disposable policy layer and a durable state layer, so an agent can rewrite its own prompts, tools and executor code without being able to damage its own event log, secrets or history.

agents · agent-harness · architecture · sandboxing · self-improvement

OpenAI hands its offensive cyber models to sixteen security firms

2026-08-15

OpenAI expanded its Daybreak Cyber Partner Program to sixteen named companies including Accenture, IBM, Cisco, CrowdStrike and Cloudflare, letting them embed its frontier cyber models in their own products while keeping model access away from end customers.

cybersecurity · ai-security · red-teaming · openai · vulnerabilities · model-access

A fired xAI engineer says he was cut days before presenting safety findings

2026-08-15

A wrongful-termination complaint filed in Santa Clara County alleges an early xAI engineer was fired shortly before presenting AI-safety findings to leadership, and it sits against a verified record of a Canadian regulator ruling that Grok's image tool launched without proper safeguards.

cybersecurity · ai-safety · privacy · regulation · xai · whistleblower

Qwen passed one billion downloads, not three billion

2026-08-15

Alibaba Cloud's own announcement says its Qwen models crossed 1 billion cumulative Hugging Face downloads at about 1.1 million a day with 200,000 derivative models, passing Meta's Llama, contradicting a widely repeated claim of 3 billion downloads past both Meta and Google.

open-weight-models · qwen · alibaba · distribution · policy · china

The open world model ships inference and keeps the training code

2026-08-15

AlayaWorld released inference code and pretrained weights for an interactive world model with long-horizon memory, but the training code is still an unchecked box, the license is a community license, and running it requires a gated Google model plus a ByteDance depth model.

world-models · video-generation · open-weights · licensing · local-ai

One layer creates the giant activations behind attention sinks

2026-08-15

Researchers identified a single layer, consistent across model families, where the outsized activations that produce attention sinks first appear, and showed that loosening that token's rigidity improves instruction following and math reasoning without retraining.

interpretability · mechanistic-interpretability · attention · efficiency · research

← 2026-08-142026-08-15later →