yoppi

I'm a software engineer and runner.

[yoppi @yoppiblog]

https://t.co/sOPjRkQbM3 僕も解きましたがとてもいい問題です。一緒にML・LLMやっていく方お待ちしております。

[Noam Brown @polynoamial]

After AlphaGo, the skill of human Go players noticeably improved. I suspect we will see a similar pattern in math.

[yoppi @yoppiblog]

MarkdownからHTMLか〜。今までもXMLタグは有効だったし、統一されるといいな

[Sakana AI @SakanaAILabs]

What if instead of building one giant AI, we evolved a coordinator to orchestrate a diverse team of specialized AIs? 🐟

Excited to share our new paper: “TRINITY: An Evolved LLM Coordinator”, published as a conference paper at #ICLR2026!

Paper: https://t.co/Pr1TT15iqa

In nature, complex problems are rarely solved by a single monolithic entity, but rather by the coordinated efforts of specialized individuals working together. Yet, modern AI development is heavily focused on endlessly scaling up single, massive monolithic models, yielding diminishing returns. While model merging offers a way to combine different skills, it is often impractical due to mismatched neural architectures and the closed-source nature of top-performing models.

To address this, we took a macro-level approach: test-time model composition. We introduce TRINITY, a system that fuses the complementary strengths of diverse, state-of-the-art models without needing to modify their underlying weights.

TRINITY processes queries over multiple turns. At each step, a lightweight coordinator assigns one of three distinct roles to an LLM from its available pool:

1/ Thinker: Devises high-level strategies and analyzes the current state.
2/ Worker: Executes concrete problem-solving steps.
3/ Verifier: Evaluates if the current solution is complete and correct.

By dynamically assigning these roles, the coordinator effectively offloads complex reasoning and skill execution onto the external models.

What makes TRINITY unique is its extreme efficiency. The coordinator relies on the hidden states of a compact language model and a small routing head. In total, it has fewer than 20K learnable parameters.

Training this system presented a massive challenge. Traditional Reinforcement Learning (REINFORCE) failed because the gradients had a low signal-to-noise ratio due to binary rewards and weak parameter coupling. Imitation learning (Supervised Fine-Tuning) was ruled out because generating multi-turn labels is prohibitively expensive.

Our solution? We turned to nature-inspired algorithms. We optimized the coordinator using a derivative-free evolutionary algorithm. We found that evolution is uniquely suited to optimize this tight, high-dimensional coordination problem where traditional gradient-based methods fail.

The results are very promising. In our experiments, TRINITY consistently outperforms existing multi-agent methods and individual models across various benchmarks. At the time of publication, it set a new state-of-the-art record on LiveCodeBench, achieving an 86.2% pass@1 score.

More importantly, it demonstrated incredible generalization. Without any retraining, TRINITY transferred zero-shot to four unseen tasks (AIME, BigCodeBench, MT-Bench, and GPQA). On average, the evolved coordinator surpassed every individual constituent model in its pool, including GPT-5, Gemini 2.5-Pro, and Claude-4-Sonnet (the top frontier models available at the time of our #ICLR2026 submission last year).

This work is central to Sakana AI's vision. We believe the future of AI isn't just about scaling monolithic models, but engineering collaborative, diverse AI ecosystems that can adapt and combine their strengths.

We invite the community to read the paper and explore these ideas!

Paper: https://t.co/Pr1TT15iqa
OpenReview: https://t.co/mbSP1aFCv7

This foundational research is part of the core engine powering our multi-agent product: Sakana Fugu 🐡👇

[yoppi @yoppiblog]

大峯奥駈道、また走るぞ〜。2泊3日計画

[yoppi @yoppiblog]

https://t.co/0vsz94HaAx git-pr-release、好きすぎてずっとプロジェクトで使っていましたが、Rubyを離れたいのでRustでportしました

[yoppi @yoppiblog]

ちょっとRustわかるようになってきた

[yoppi @yoppiblog]

@tapioca_pudd Composer 2 いいですよね（Kimi K 2.5シンパ）

[yoppi @yoppiblog]

競プロの問題解いてみたら驚くほど書けなくなってしまって悲しい...

[yoppi @yoppiblog]

CCのChannels、IRCのLimeChatっぽいことができるな（結局そこに戻ってくる感）

[yoppi @yoppiblog]

ぎわさんにもいなむらさんにも会ったしNLPはゆるく繋がれていいところだ

[yoppi @yoppiblog]

最近の
* 寝る前にタスク整理して、Agent Teamsに依頼して投げて寝る
* 朝起きて確認してTODOタスクをVimプラグインで整理
* 粛々と繰り返して、1つめ戻る

[yoppi @yoppiblog]

AI時代に訓練されたtmuxさばきが活きるとは

[yoppi @yoppiblog]

neovim移行してみたけど、結局MacVimから逃れられない...（細やかな打鍵感の違い）

[yoppi @yoppiblog]

今、基盤モデルにアクセスできているけど、アクセスを禁じられたら困惑するくらいには、すでに、依存してしきっている

[Takuya Akiba @iwiwi]

巨大なLLM事前学習データを爆速で検索出来る「SoftMatcha 2」の開発に参加させてもらいました。デモ、論文、ソースコード等をこの度公開しましたので是非お試し下さい！

https://t.co/hVDXc60F5C

意味的類似性に基づいた置換や挿入削除に対応しながら1兆トークン規模のデータを0.1秒代で検索すると いうなかなか狂った性能になってます。EMNLP'25 Best Paperのinfini-gram-miniを含む既存のツール全てを大きく凌駕する性能だと思います。用途に特化したデータレイアウトを持つdisk-aware suffix arrayを使いながら、本来指数的になる置換・挿入・削除の候補を実データに基づきうまく枝刈りすることで高速な検索を達成してます。

この規模の事前学習データを検索出来ることの利点の事例として、論文ではベンチマークの汚染の検証をやってみてます。infini-gram-miniのような厳密な検索のみでは発見出来ないような汚染の事例なども有りそうでした。

現在デモでは数百Bトークン規模のデータからの検索を試せるようになってます。コードも公開してますのでご自身でホストしてもらうとより大規模なケースもお試し頂けます。

🌐 Demo: https://t.co/iNHhzeIsw3
📄 Paper: https://t.co/HNxk0aK5WS
💻 Code: https://t.co/foPSdoviuz

若き才能 @e869120 を始めとするSoftMatchaチームの方々との協働はとても刺激的で多くの学びがありました。楽しかった〜！ありがとうございました！ @shiatsumat @go2oo2 @ksuenaga @MasWag @sho_yokoi

[yoppi @yoppiblog]

骨が問題だと思ってたら筋膜だったらしい（ブドウ糖注射したら良くなった）。あと、ストレッチ。少し走れるようになった😭

[yoppi @yoppiblog]

足の故障で山を走れず、土日がほんとに無

[yoppi @yoppiblog]

まって、Kimi K2.5聞いてない。fine-tuningしてみたい

[yoppi @yoppiblog]

左足が負荷で壊れて一ヶ月... ADIZERO EVO SLはめっちゃいいんだけど、僕の技術ではまだ練習用で使うには早かったな...

[yoppi @yoppiblog]

https://t.co/TWtfGbTIqw
考えていることが言語化されてた