開発者ツール
B
326 件中 31〜60 件目を表示
-
Beyond Outcomes: Dual-View Relational Learning for Efficient Agent Benchmarking
-
How Much is a Human Right Worth? ECtHR-NPD: A Benchmark for Predicting Non-Pecuniary Damage Awards
-
Structured Claim-Level Discourse Representations for Dense Health Narratives
-
Fast Learning Rates for Physics-Informed Kernel Methods
-
Translating CUDA Tile Operations from Python to Rust Using Agentic AINVIDIA、AI エージェントで CUDA タイルカーネルを Rust へ全移植NVIDIA が、GPU カーネルを Rust で安全に記述するタイル基盤 cuTile Rust と、既存の cuTile Python / Triton-TileIR カーネルをそこへ翻訳する AI エージェントスキルを公開。解析・カーネル・ホスト/FFI・ベンチマークの各段を機械検証する多段パイプラインで変換し、TileGym の公開 24 演算子 (約 40 カーネル) を全移植、cuTile Python 比で平均 99.5% の性能に達したという。
-
Learning Lyapunov Operators for Nonlinear Systems
-
Preventing Model Collapse: A Fisher-Rao Perspective on the Dynamics of Training with Synthetic Data
-
ASLEval: Measuring Privacy Exposure Displacement in LLM Agent Sessions
-
Physics-based prediction, uncertainty quantification and decision-making for IN718 crystallographic texture intensity across LPBF defocus regimes
-
Quoting Mustafa SuleymanMicrosoft AI の Suleyman 氏、モデルに権利を認めるなと警告Simon Willison が、Microsoft AI CEO のムスタファ・スレイマン氏による論考「A warning about 'model welfare'」を引用した。スレイマン氏は、モデルを感情や選好、権利を持つ存在として扱うべきではないと主張。意識こそが倫理・法・政治の体系の基盤であり、他の存在にその種の権利を分け与えることは証拠に照らして正当化されず、AI の封じ込めとアラインメントをいっそう難しくすると論じている。
-
PersonaPath: Towards Knowledge-Centric Personalized Learning Path Planning
-
GrainSpeech: Less Context, More Detail for Compact Speech Synthesis
-
EviGen: Predictive Evidence Scaffolding for Verifiable Clinical Rationale Generation
-
Ask the Tool, Don't Guess: Agent Tool Calls Hold Their Progress, and the Serving System Should Read It
-
ReFigBench: Benchmarking Scientific Figure Reconstruction as Editable PowerPoint Artifacts
-
Interpretable Multi-Instance Learning Enables Early Prediction of Key Molecular Alterations from Routine Flow Cytometry in Acute Myeloid Leukemia
-
Using OCR Heads to Verbalize Image Semantics
-
ProgramDistill: From Interactive Web Apps to Verifiable Reference-Guided SWE Tasks
-
Beyond frequency measures: Can contextual embeddings capture meaning change in scientific texts?
-
Inside Sakana AI's Product TeamSakana AI、プロダクトチームの文化・働き方・採用プロセスを公開Sakana AIが研究成果を製品化するプロダクトチームの内部を紹介する記事を公開。チーム構成やコアタイム無しの柔軟な勤務形態、Applied Research EngineerとSoftware Engineerの一日、採用プロセス、求める人物像を、メンバーへのインタビューに基づきFAQ形式でまとめた。
-
Zero-Shot Cross-Lingual Recognition of Sign Language Handshapes
-
Version- and Scope-Aware Question Answering over Normative Documents: A Deployed System and an End-to-End Evaluation at Production Scale
-
TeleAntiFraud 2.0: A Refreshable, Profile-Grounded, and Audio-Based Benchmark for Telecom Fraud Detection
-
When Edit Flows are Edit Jumps: replicating Edit Flows and EvoFlows
-
A Scalable Framework for Automated NER Annotation Correction in Low-Resource Languages
-
Which LLM is Best for Translating Natural Language Goals to PDDL
-
"If I Had to Buy Just ONE: Galaxy S26 Ultra": Auditing AI-Generated Product Recommendations
-
LocQE: Principled Domain Adaptation for Localisation Quality Estimation by Leveraging Post-Edits
-
Rethinking Critic Learning in PPO: Understanding and Mitigating Value Flattening
-
Echo: Learning-based Matching Decompilation using Trusted Back Translation