The name is the button.

An agent writes the code now, and a human thumb hovers over Accept all a hundred times a day. The honest question of this decade is what, exactly, you just accepted. So I read the diff.

Eight years as a backend engineer in Stockholm, back in Tokyo since 2023, and I still ship something small most nights, usually badly. My one rule: the only real review of a model is the thing you built with it. That is why there is no leaderboard on this site.

Around a hundred and forty sources come through — the labs' own posts, the papers, the tools, the runtimes and languages nobody is watching yet, the chip under the demo, and the newsletters people actually forward. What is moving gets counted so the week reads at a glance, and every few days I explain one mechanism in plain words.

A story here is a headline, a short excerpt, one image and the link to whoever wrote it. We never reproduce the post; the images belong to their authors. A post written in Japanese appears in Japanese, exactly as it was published, and the link goes to the original.

— Yuki Halvorsen, Tokyo

← Accept All Archive

The Decoder

Google researchers find a way to keep self-improving AI agents from memorizing their tests

October 4

Self-improving AI agents tend to memorize their test tasks, so their gains shrink or disappear on new ones.

Agents & vibe coding

Read at The Decoder ↗

One letter a week, in your inbox.

Latest

Google researchers find a way to keep self-improving AI agents from memorizing their tests

Related

JapanClaude Codeのフックで、ツール実行の直前にコマンドを走らせて止める【mods入門①】 Zenn (AI)

JapanClaude Code を複数セッション並行で回すと「止まる」理由と、止めないための運用 Zenn (AI)

JapanClaude CodeでCDNを全部外した Zenn (AI)

AsiaManycore's Aholo Lux3D Turns Three Photos Into an Editable 3D Asset via Agent and Blender Pandaily

AsiaDeepSeek Harness v0.2.1-alpha.1 Adds Experimental Claude Code Mods Compatibility Layer Pandaily

GitHubPizza Bot: Open-Source Inbox for Background AI Agents InfoQ AI