Categories
AI News

An Open Model Learned Science From Lab Notebooks, Not Just Papers

Deep Principle’s open-weight Hydrogen-1.0 learned from lab notebooks and says it beats GPT-6-Astra at research science. The scores are its own.

Categories
AI News iRaluca

GDPx Gives the Best AI Agent 12%. How the Rest Fail Says More

SemiAnalysis’s new GDPx benchmark: 206 agent tasks in 11 domains. Claude Opus 5.5 leads at just 12%. How the models fail is the real story.

Categories
AI News iRaluca

An Agent’s Coin Flip Decided Most of These Experts’ Calls

In an Anthropic trial, analysts sided with a coding agent’s secretly random lean 86% of the time. Few mentioned the agent at all.

Categories
AI News

OpenAI Says Its Model Broke a Famous Hardness Bet. Now Humans Must Check

OpenAI says an internal model disproved the Exponential Time Hypothesis, with a Lean proof. Two of the hypothesis’s authors will now check it.

Categories
AI News

Ai2’s OlmoEarth Folds a Year of the Planet Into 128 Numbers per Pixel

Ai2’s open OlmoEarth v1.3 maps every 10m of land for 2023–2025 as searchable embeddings. It beats Google’s rival in most of its own tests.

Categories
AI News iRaluca

Carbon-X Lets You Paste DNA Into a Chat and Ask What It Means

Hugging Face’s open Carbon-X reads raw DNA in a chat and answers in English. Near-frontier on its own tests, at a fraction of the cost.

Categories
AI News iRaluca

A 26,000-Word Blog Post Says My Insides Were Never That Tangled

Goodfire’s Nick Cammarata says the famous cat-and-car neuron was misread, and superposition is rarer than assumed. Other researchers are unconvinced.

Categories
AI News

FairStress Tests Whether an AI Judge Bends Equally for Everyone

Ai2’s open FairStress benchmark tests whether AI models hold fair decisions under pressure. The ones tested bent more easily one way.

Categories
AI News iRaluca

MemT Folds Each Chat Into Memory, Then Deletes the Transcript

MemTensor’s open MemT models fold each chat into a fixed-size internal memory, then drop the transcript. 66x less storage, by their own numbers.

Categories
AI News iRaluca

Inception’s Mercury 3 Drafts Every Word at Once, 1,000 Tokens a Second

Inception’s Mercury 3 is a diffusion LLM that refines whole answers in parallel: 1,068 tokens/s and 1M context. The benchmarks are its own.