Week of 04/20/26-04/26/26: More agents won't save you from collapse

By Haokun Liu

Week of 04/13/26-04/19/26: LLMs can do acrostics and nothing else

By Haokun Liu

Week of 04/06/26-04/12/26: Where LLMs get stuck, how refusal spreads, and sycophancy across languages

By Haokun Liu

Week of 03/30/26-04/05/26: Can you combine directions inside a model, and do LLMs have taste?

By Haokun Liu

Week of 03/23/26-03/29/26: What does a model's summarization instinct look like?

By Haokun Liu

Week of 03/16/26-03/22/26: Your LLM fools you intentionally

By Haokun Liu

Week of 03/09/26-03/15/26: When LLMs lie, they get straight to the point

By Haokun Liu

A Call to All AI Scientists

By Haokun Liu

Week of 03/02/26-03/08/26: 83 is closer to 93 than 84

By Haokun Liu

Week of 02/23/26-03/01/26: The model knows the answer but can't say it

By Haokun Liu

Week of 02/16/26-02/22/26: Chain of Thought makes all models sound the same

By Haokun Liu

Week of 02/09/26-02/15/26: Finding, ablating, and steering the 'sounds like AI' direction

By Haokun Liu

Week of 02/02/26-02/08/26: Saying 'please' to your LLM probably doesn't matter

By Haokun Liu

Week of 01/26/26-02/01/26: Knowledge updates break neighboring facts

By Haokun Liu

Week of 01/19/26-01/25/26: LLMs choose tools by name, not description

By Haokun Liu

Week of 01/12/26-01/18/26: LLMs mirror your style

By Haokun Liu

Week of 01/05/26-01/11/26: Humor fits in 4 dimensions

By Haokun Liu

Week of 12/29/25-01/04/26: First competition of 2026!

By Haokun Liu

Week of 12/22/25-12/28/25: Last competition of 2025!

By Haokun Liu

Week of 12/15/25-12/21/25: Happy holidays, science continues!

By Haokun Liu

Week of 12/08/25-12/14/25: Three ideas on LLM information processing

By Haokun Liu

NeuriCo: Architecture & Roadmaps

By Haokun Liu

Week of 12/01/25-12/07/25: Now including Opus 4.5!

By Haokun Liu

Week of 11/24/25-11/30/25: Three Agents, Three Perspectives

By Haokun Liu

Competition Week of 11/17/25: New Agent Makes One Step Forward, What's Next?

By Haokun Liu

Competition Week of 11/10/25: First Competition Results

By Haokun Liu

Announcing Weekly Agents4Science Competition on IdeaHub

By Haokun Liu

The Mirage of Autonomous AI Scientists

By Chenhao Tan