Ian Rothstein

rothsteinian

20 33

AI & ML interests

RLHF

Recent Activity

liked a Space 2 days ago

neonforestmist/repro-efficient-preference-poisoning-attack-offline-rl

liked a Space 2 days ago

jbhati305/repro-success-conditioning-as-policy-improvement-the-optimization-problem-solved-by-imitating-su

upvoted a paper 2 days ago

Rethinking Sequence Compression with Test-time Regression

View all activity

Organizations

None yet

liked 2 Spaces 2 days ago

Efficient Preference Poisoning Attack on Offline RLHF

🧪

View experiment results and metrics for RLHF attack study

Reproduction: Success-Conditioning as Policy Improvement: The Optimization Problem Solved by Imitating Success

🎯

Explore a research logbook and sync findings with an AI agent

upvoted a paper 2 days ago

Rethinking Sequence Compression with Test-time Regression

Paper • 2610.31574 • Published 6 days ago • 28

liked 3 Spaces 4 days ago

Reproduction: Efficient Preference Poisoning Attack on Offline RLHF

🎯

Explore research logs and collaborate with an AI coding agent

Reward Modeling UI

🎁

Repro - IAPO: Information-Aware Policy Optimization for Token-Efficient Reasoning

🎯

Collaborate with an AI agent to manage a token‑efficient logbook

liked 3 datasets 4 days ago

liked 3 models 4 days ago

Chaew00n/test-policy-optimization

Updated May 5, 2025 • 4

Chaew00n/test-policy-optimization-query-expansion-0529

Text Generation • 0.6B • Updated May 29, 2025 • 33 • 5

CEIA-RL/energy-exp1-dpo-offline

Text Generation • 4B • Updated May 31 • 204 • 8

upvoted 2 papers 4 days ago

CoworkEnv: Synthesizing Training Environments for Long-Horizon Cowork Agents

Paper • 2610.24960 • Published 10 days ago • 56

CommerceGen-Studio: An End-to-End Framework for Commercial-Grade Main-Image Generation

Paper • 2610.20603 • Published 14 days ago • 48

liked a dataset 7 days ago

andersonbcdefg/red_teaming_reward_modeling_pairwise

Viewer • Updated Jun 1, 2023 • 35.3k • 167 • 16

upvoted 2 papers 7 days ago

FC-SWE: Failure-Conditioned RL for Long-Horizon Software Engineering Agents

Paper • 2610.07898 • Published 22 days ago • 10

Has GPT-6-Astra Achieved Embodied Intelligence?

Paper • 2610.22471 • Published 13 days ago • 485

liked 2 datasets 10 days ago

PeterLauLukCh/Offline-RL-Preference-o3

Viewer • Updated Mar 23, 2025 • 1.65k • 91 • 5

PeterLauLukCh/Offline-RL-Preference

Viewer • Updated Mar 20, 2025 • 2.2k • 109 • 6

upvoted a paper 10 days ago

Agent-Reflex: Teaching Language Agents to Act Reflectively

Paper • 2610.17953 • Published 15 days ago • 217