Log inSign up
Hiroki Furuta
160 posts
user avatar
Hiroki Furuta
@frt03_
Research Scientist at @GoogleDeepMind / PhD from The University of Tokyo @Matsuo_Lab / AI Agent / AI Alignment / LLMs / Deep RL
frt03.github.io
Joined August 2018
354
Following
1,016
Followers
  • user avatar
    Hiroki Furuta
    @frt03_
    Sep 26, 2024
    RLHFに関する論文が #NeurIPS 2024に採択されました!🎉 DPOが好みの割合を表すソフトな選好ラベルを活用できるように拡張し、オフラインとオンラインの両方でより良いLLMのアラインメントを達成します Google DeepMindでのインターン中の成果です
    user avatar
    Hiroki Furuta
    @frt03_
    Sep 26, 2024
    Our paper on RLHF: Geometric-Averaged Preference Optimization for Soft Preference Labels was accepted to #NeurIPS 2024! To deal with over-optimization in DPO, proportional soft labels taken from majority voting/AI feedback can adjust the gradient scale. arxiv.org/abs/2409.06691
    Image
    Image
    Image
    42K
  • user avatar
    Hiroki Furuta
    @frt03_
    Jan 23, 2025
    Our paper about the evaluation of sparse autoencoder (SAE) was accepted at #ICLR2025 🎉 We propose leveraging the representation similarity of polysemous words (e.g. "space") to measure the quality of SAE. This is work done at The Univerisity of Tokyo @Matsuo_Lab
    Image
    28K
  • user avatar
    Hiroki Furuta
    @frt03_
    Jan 23, 2025
    東大松尾・岩澤研究室で行った共著の論文が #ICLR 2025に採択されました🎉 GPT-4/LLaMA/Gemmaなどにも用いられている、LLMの内部回路の解釈性(mechanistic interpretability)を得るためのツールであるSAEの性能を、多義語を用いて測る新たな指標を提案しました
    user avatar
    Hiroki Furuta
    @frt03_
    Jan 23, 2025
    Our paper about the evaluation of sparse autoencoder (SAE) was accepted at #ICLR2025 🎉 We propose leveraging the representation similarity of polysemous words (e.g. "space") to measure the quality of SAE. This is work done at The Univerisity of Tokyo @Matsuo_Lab
    Image
    17K
  • user avatar
    Hiroki Furuta
    @frt03_
    Sep 29, 2021
    既存の深層強化学習のアルゴリズムを分類した論文がNeurIPS2021に採択されました!松尾研、@Tdash_Kozさん、 @shaneguMLさんの共同研究の成果です arxiv.org/abs/2103.17258
  • user avatar
    Hiroki Furuta
    @frt03_
    Jan 24, 2023
    "A System for Morphology-Task Generalization via Unified Representation and Behavior Distillation" was accepted at #ICLR2023 as notable-top-25% (w/ @yusuke_iwasawa_ @ymatsuo @shaneguML) We propose a data distillation pipeline for large-scale models in RL. arxiv.org/abs/2211.14296
    Image
    46K
  • user avatar
    Hiroki Furuta
    @frt03_
    Feb 5, 2025
    Our new preprint on inference-time alignment for diffusion models is out🎉 To improve text-to-video generation, we propose Diffusion Latent Beam Search with Lookahead Estimator. We should allocate the computation as: Lookahead Steps >> Search Budget >> Denoising Steps 1/8
    Image
    GIF
    18K
  • user avatar
    Hiroki Furuta
    @frt03_
    May 25, 2023
    Yesterday I droped by AK meetup in Tokyo and have dinner with him! Really exciting night🤗 Enjoy your rest of stay in Japan @_akhaliq
    Image
    32K
  • user avatar
    Hiroki Furuta
    @frt03_
    Dec 6, 2024
    Google DeepMind Japanでのインターン中の成果をプレプリントで公開しました! 物体がダイナミックに動くような場面の生成では不正確な出力が多いtext-to-videoの動画生成モデルを、GeminiなどマルチモーダルなVLMからのAI feedbackによるRL finetuningを通して改善します
    user avatar
    Hiroki Furuta
    @frt03_
    Dec 5, 2024
    Text-to-video models can generate photorealistic scenes but still struggle to accurately depict dynamic object interactions😢 Our new preprint addresses this through RL finetuning with AI feedback from VLMs capable of video understanding (e.g. Gemini, etc)🎉 1/7
    Image
    GIF
    13K
  • user avatar
    Hiroki Furuta
    @frt03_
    May 8, 2021
    強化学習のベンチマーク環境の難易度を定量評価するための指標(PIC/POIC)を提案した論文がICML2021に採択されました! 論文: arxiv.org/abs/2103.12726 解説資料(日本語):
    user avatar
    DLHacks
    @DL_Hacks
    May 6, 2021
    PIC/POICは報酬/最適性変数と方策のパラメータ間の相互情報量によって定義され、強化学習ベンチマーク環境の"解きやすさ"を特定のアルゴリズムに依らず定量評価できる。また学習せずにいくつかの実験パラメータが調整できる。 slideshare.net/DeepLearningJP…
    arXiv logo
    arxiv.org
    Policy Information Capacity: Information-Theoretic Measure for...
    Progress in deep reinforcement learning (RL) research is largely enabled by benchmark task environments. However, analyzing the nature of those environments is often overlooked. In particular, we...
  • user avatar
    Hiroki Furuta
    @frt03_
    Aug 24, 2023
    FORBES JAPAN 30 UNDER 30 2023に選んで頂きました。これからも良い研究ができるように頑張ります! #u30fj
    Image
    FORBES JAPAN 30 UNDER 30 2023|日本発「世界を変える30歳未満」120人
    From forbesjapan.com
    8.9K
  • user avatar
    Hiroki Furuta
    @frt03_
    Sep 26, 2024
    Our paper on RLHF: Geometric-Averaged Preference Optimization for Soft Preference Labels was accepted to #NeurIPS 2024! To deal with over-optimization in DPO, proportional soft labels taken from majority voting/AI feedback can adjust the gradient scale. arxiv.org/abs/2409.06691
    Image
    Image
    Image
    35K
  • user avatar
    Hiroki Furuta
    @frt03_
    Dec 5, 2024
    Text-to-video models can generate photorealistic scenes but still struggle to accurately depict dynamic object interactions😢 Our new preprint addresses this through RL finetuning with AI feedback from VLMs capable of video understanding (e.g. Gemini, etc)🎉 1/7
    Image
    GIF
    27K
  • user avatar
    Hiroki Furuta
    @frt03_
    Jan 16, 2024
    Two papers on web agents are accepted at #ICLR2024 (w/ Oral)! Thank you to all the collaborators. [Poster] Multimodal Web Navigation arxiv.org/abs/2305.11854 [Oral] A Real-World WebAgent with Planning, Long Context Understanding, and Program Synthesis arxiv.org/abs/2307.12856
    arXiv logo
    arxiv.org
    Multimodal Web Navigation with Instruction-Finetuned Foundation Models
    The progress of autonomous web navigation has been hindered by the dependence on billions of exploratory interactions via online reinforcement learning, and domain-specific model designs that make...
    5.1K
  • user avatar
    Hiroki Furuta
    @frt03_
    Feb 5, 2025
    拡散モデルのInference-time Alignmentに関するPreprintを公開しました🎉 各denoising stepで、最終的なデータ点の報酬評価に基づく潜在変数のビームサーチを行い、より好ましい動画をtext-to-videoで生成します。計算コストに対するスケーリングが見られます。 東大 松尾・岩澤研での研究成果です
    user avatar
    Hiroki Furuta
    @frt03_
    Feb 5, 2025
    Our new preprint on inference-time alignment for diffusion models is out🎉 To improve text-to-video generation, we propose Diffusion Latent Beam Search with Lookahead Estimator. We should allocate the computation as: Lookahead Steps >> Search Budget >> Denoising Steps 1/8
    Image
    GIF
    8.6K

New to X?

Sign up now to get your own personalized timeline!

Create account

By signing up, you agree to the Terms of Service and Privacy Policy, including Cookie Use.

Terms·Privacy·Cookies·Accessibility·Ads Info·© 2026 X Corp.
Don't miss what's happening
People on X are the first to know.
Log inSign up
Advertisement
Advertisement