Researcher — paper.dinq.me

2026 顶会论文

作者位次 #4/5NeurIPS 2026Poster

LLMs Keep Thinking When Told Not To

Dianqiao Lei · Kevin Qinghong Lin · Pan Lu · Philip Torr · James Zou

作者位次 #13/18NeurIPS 2026Poster

SETA: Scaling Environments for Terminal Agents

Qijia Shen · Zhiqi Huang · Vamsidhar Kamanuru · Aznaur Aliev · Jay Rainton · Ahmed Awelkair · Boyuan Ma · Qizheng Zhang · Jiwei Fu · Yuzhen Mao · Wendong Fan · Ping Nie · Philip Torr · Bernard Ghanem · Changran Hu · Jonathan Li · Urmish Thakker · Guohao Li

智能体与工具OpenReview

作者位次 #8/9NeurIPS 2026Poster

Inter-Agent Influence: Evaluating Persuasion, Deception and Coercion in Multi-Agent Systems

Chandler Smith · Cecilia E Tilli · Qi Guo · Sophia Hatz · David Demitri Africa · Patricia Paskov · Christian Schroeder de Witt · Philip Torr · Lewis Hammond

智能体与工具评测与基准社会与人本OpenReview

作者位次 #10/16NeurIPS 2026Poster

Evaluating AI-based Scientific Knowledge Synthesis with Epidemiological Systematic Reviews

Shreyansh Padarha · Ryan Othniel Kearns · Tristan M Naidoo · Lingyi Yang · Łukasz Borchmann · Piotr Blaszczyk · Christian Morgenstern · Ruth McCabe · Sangeeta Bhatia · Philip Torr · Jakob Foerster · Scott Hale · Thomas Rawson · Anne Cori · Elizaveta Semenova · Adam Mahdi

推理评测与基准检索与知识OpenReview

作者位次 #2/3NeurIPS 2026Poster

Tight PAC-Bayes Generalisation Guarantees for Large Language Model Safety Monitoring

Tom Lamb · Philip Torr · Tim G. J. Rudner

安全OpenReview

作者位次 #5/6NeurIPS 2026Poster· Social Aspects / Fairness, Equity, Justice and Safety

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer

Yupeng Chen · Junchi Yu · Aoxi Liu · Baoyuan Wu · Philip Torr · Adel Bibi

安全社会与人本对齐与偏好OpenReview

作者位次 #4/6NeurIPS 2026Poster

ActO: Extracting Action Representations from MLLM Embeddings for Video World Models

Runjia Li · Minghao Chen · Junyu Xie · Philip Torr · Andrea Vedaldi · Tomas Jakab

可解释性多模态OpenReview

作者位次 #4/7NeurIPS 2026Poster· Social Aspects / Accountability, Transparency and Interpretability

MMDiff: Multimodal Model Diffing for Feature Discovery and Control

Lachin Naghashyar · Hunar Batra · Ashkan Khakzar · Philip Torr · Ronald Clark · Christian Schroeder de Witt · Constantin Venhoff

可解释性多模态社会与人本OpenReview

作者位次 #3/3NeurIPS 2026Poster

How LLMs Distinguish Threats from Offers

Julia Karbing · Lewis Hammond · Philip Torr

OpenReview

作者位次 #5/8NeurIPS 2026Poster

StraTA: Incentivizing Agentic Reinforcement Learning with Strategic Trajectory Abstraction

Xiangyuan Xue · Yifan Zhou · ZiDong Wang · Shengji Tang · Philip Torr · Wanli Ouyang · LEI BAI · Zhenfei Yin

智能体与工具强化学习与后训练OpenReview

Philip Torr

获取完整研究档案

联系方式