← Back to feed

"About a quarter of them contained bugs": Gemini 4's ML R&D benchmark

L5 · ResearcherResearchLessWrong AI· 10/16/2026

Critical analysis of benchmark methodology flaws in major AI safety research directly impacts ML researchers evaluating model capabilities.

AI Summary

Google DeepMind's Gemini 4 safety report reveals 25% of its ML R&D tasks contained bugs, undermining claims about AI automating AI research.

Excerpt

Google DeepMind’s Gemini 4 Argon safety report[1] rules out its “AI automating AI research” thresholds (the ones most relevant to RSI) using a benchmark that, by the report’s own audit, has about a quarter of its scored tasks unpassable. This post dives a bit deeper into the context around this disturbing line tucked away on page 67 of the report. This is no knock on the incredible Google engineers and scientists who worked tirelessly to test Gemini as thoroughly as they could and generate this

Read Original
0 upvotes · 0 downvotes · 1 min read

Related Articles

L5 · ResearcherResearchHacker News
Claude reveals ECDSA curve origins

Claude AI discovered that NIST's ECDSA cryptographic seeds were generated from Unix timestamps using lrand48 RNG, solving a decades-old mystery in cryptography.

L4 · DeveloperResearchHacker News
90% of US working hours are spent on tasks not covered well by AI benchmarks

Mercor Research found that 90% of US working hours involve tasks with little to no coverage in existing AI benchmarks, making it difficult to assess model performance on real-world work.

L5 · ResearcherResearchHacker News
Neural networks are clean

New research challenges the superposition theory of polysemanticity, suggesting neural networks are more interpretable than previously believed through recursive generalization patterns.

L5 · ResearcherResearchAI Alignment Forum
A Simpler and Better Debate Protocol for Scalable Oversight

Researchers propose a new AI debate protocol that offers simpler implementation and stronger correctness guarantees for scalable oversight of advanced AI systems.

L4 · DeveloperResearchMIT Technology Review AI
Roundtables: A Conversation With the Creator of AI-Designed Viruses

MIT Technology Review interviews Stanford PhD student Samuel King about using generative AI to design viral blueprints, exploring AI's potential for creating biological systems.