技术雷达 · 2026-09-06

G GitHub 新星 10 条

A arXiv 论文 10 条

1 Compile by Training: Turning Natural-Language Specifications into Local Neural Functions cs.CLcs.AIcs.LG
Many recurring text functions are easy to describe but difficult to implement with rules, while calling a large remote model for every input introduces repeated cost, latency, and dependency on a prov
Yuntian Deng, Pengyu Nie, Stuart Shieber
2 Clean Engineering, Unstable Measurement: A Preregistered Reliability Failure of Black-Box LLM Observers on Shared Endpoints cs.AIcs.LG
Language-model judges now gate training data, score generations, and drive leaderboards. The judge is then a measurement instrument, resting on one rarely stated assumption: the same request, sent to
Haoyaun Zhu, Jie Zhang
3 ESPO: Error-Structured Prompt Optimization via Diagnose, Diversify, and Stabilize cs.CLcs.AI
Evolutionary prompt optimizers such as GEPA suffer from prompt bloat: each iteration appends rules and caveats, producing prompts up to 3$\times$ longer yet no more accurate. We trace this to three de
Lihao Liu, Peng Tang, Kunwar Yashraj Singh
4 Legibility is Not Interpretability: Comparing Judged and Actual Importance in Chain-Of-Thought Reasoning cs.CLcs.LG
Reasoning traces from chain-of-thought models appear to offer a legible window into how a model arrives at its answer. A growing body of work treats them as such, using LLM judges to diagnose errors,
Kevin Du, Alexander Hoyle, Laura Ruis
5 One Editor, Many Edits: A Unified Training-Free Framework for Diverse Video Editing cs.CVcs.AI
Video editing spans diverse editing paradigms, yet achieving high-quality instruction-guided and subject-guided editing within a single unified framework remains challenging. We introduce EditVid, a t
Adheesh Sunil Juvekar, Onkar Kishor Susladkar, Kiet A. Nguyen
6 Robust PAC Learning of Concurrent Stochastic Games cs.LGcs.GTcs.LO
We introduce the first Probably Approximately Correct (PAC) learning framework for general-sum concurrent stochastic games (CSGs) with transition uncertainty, while addressing the challenge of Nash eq
Angel Y. He, David Parker
7 Seeing Before Synthesizing: VLM-Guided Transition Event Discovery for Weakly-Supervised Dense Video Captioning cs.CVcs.AI
Weakly-Supervised Dense Video Captioning aims to localize and describe multiple events in untrimmed videos given only an ordered set of event-level captions per video. Recent work synthesizes auxiliar
Ye-Chan Kim, Seunghee Choi, SeungJu Cha
8 Knowledge Acquisition During Pre-training? Large Language Models Learn Better With Auxiliary Views cs.CLcs.AI
Gaps remain in our understanding of how large language models (LLMs) acquire knowledge during pre-training. We posit that auxiliary views, reformulations of knowledge, are causally helpful for learnin
Joseph Lee, Yidi Huang, Dokyoon Kim
9 A Computationally Feasible Framework for Causal Probabilistic Explanation cs.AI
Explaining why a specific outcome occurred, and which inputs deserve the blame or credit, is central to philosophical, scientific, and policy analysis. Existing tools split into two camps. The theory
Rafal Urbaniak, Sam Witty, Daniel Waxman
10 Rethinking On-Policy Distillation of Large Language Models II: One Training Example cs.AIcs.CL
On-policy distillation (OPD) combines student-generated rollouts with dense token-level supervision from a teacher. Existing work has mainly studied its algorithmic behavior, leaving the role of train
Zixuan Fu, Bingxiang He, Yuxin Zuo

H Hacker News 10 条