Skip to content

OpenTrain Research Tools

Human Feedback and Eval Paper Explorer

A focused feed for RLHF, preference data, rater protocols, agent evaluation, and LLM-as-judge research. Every paper includes structured metadata for quick triage.

Total papers: 304 Search mode: keyword RSS

Filter by tag

All Automatic Metrics (978) General (590) Coding (314) Simulation Env (115) Math (103) Multilingual (99) Long Horizon (82) Medicine (78) Pairwise Preference (70) Law (45) Multi Agent (41) Human Eval (38) Expert Verification (25) Web Browsing (22) Critique Edit (21) Red Team (21)
Toward Safe and Human-Aligned Game Conversational Recommendation via Multi-Agent Decomposition

Zheng Hui, Xiaokai Wei, Yexi Jiang, Kevin Gao, Chen Wang, Frank Ong · Apr 26, 2025

Citations: 0
Pairwise Preference Automatic Metrics Multi Agent General
  • These domains typically involve fixed content and passive consumption, where user preferences can be matched by genre or theme.
  • In contrast, games present distinct challenges: fast-evolving catalogs, interaction-driven preferences (e.g., skill level, mechanics, hardware), and increased risk of unsafe responses in open-ended conversation.
Evaluating the Diversity and Quality of LLM Generated Content

Alexander Shypula, Shuo Li, Botong Zhang, Vishakh Padmakumar, Kayo Yin, Osbert Bastani · Apr 16, 2025

Citations: 0
Pairwise Preference Automatic Metrics General
  • Recent work suggests that preference-tuning techniques -- such as Reinforcement Learning from Human Feedback (RLHF) methods like PPO and GRPO, as well as alternatives like DPO -- reduce diversity, creating a dilemma given that these models
  • Using open-ended tasks that require no human intervention, we find counterintuitive results: when using diversity metrics that do not explicitly consider quality, preference-tuned models -- particularly those trained via RL -- often produce
Diffusion Generative Recommendation with Continuous Tokens

Haohao Qu, Shanru Lin, Yujuan Ding, Yiqi Wang, Wenqi Fan · Apr 16, 2025

Citations: 0
Pairwise Preference Automatic Metrics Coding
  • Specifically, ContRec consists of two key modules: a sigma-VAE Tokenizer, which encodes users/items with continuous tokens; and a Dispersive Diffusion module, which captures implicit user preference.
  • By conditioning on the previously generated tokens of the LLM backbone during user modeling, the Dispersive Diffusion module performs a conditional diffusion process with a novel Dispersive Loss, enabling high-quality user preference genera
Beyond Single-Turn: A Survey on Multi-Turn Interactions with Large Language Models

Yubo Li, Xiaobin Shen, Xinyu Yao, Xueying Ding, Yidi Miao, Ramayya Krishnan · Apr 7, 2025

Citations: 0
Red Team Automatic Metrics Math
  • We organize existing benchmarks and datasets into coherent categories reflecting the evolving landscape of multi-turn dialogue evaluation, and review a broad spectrum of enhancement methodologies, including model-centric strategies (in-cont
Overcoming Sparsity Artifacts in Crosscoders to Interpret Chat-Tuning

Julian Minder, Clément Dumas, Caden Juang, Bilal Chugtai, Neel Nanda · Apr 3, 2025

Citations: 0
Pairwise Preference Automatic Metrics Coding
  • Using the BatchTopK crosscoder, we successfully identify a set of chat-specific latents that are both interpretable and causally effective, representing concepts such as $\textit{false information}$ and $\textit{personal question}$, along w
A Scalable Framework for Evaluating Health Language Models

Neil Mallinar, A. Ali Heydari, Xin Liu, Anthony Z. Faranesh, Brent Winslow, Nova Hammerquist · Mar 30, 2025

Citations: 0
Rubric RatingExpert Verification Automatic Metrics Medicine
  • As LLM-driven health applications are increasingly adopted, rigorous and efficient one-sided evaluation methodologies are crucial to ensure response quality across multiple dimensions, including accuracy, personalization and safety.
  • Current evaluation practices for open-ended text responses heavily rely on human experts.
MedPlan: A Two-Stage RAG-Based System for Personalized Medical Plan Generation

Hsin-Ling Hsu, Cong-Tinh Dao, Luning Wang, Zitao Shuai, Thao Nguyen Minh Phan, Jun-En Ding · Mar 23, 2025

Citations: 0
Expert Verification Automatic Metrics Medicine
  • Comprehensive evaluation demonstrates that our method significantly outperforms baseline approaches in both assessment accuracy and treatment plan quality.
Measuring AI Ability to Complete Long Software Tasks

Thomas Kwa, Ben West, Joel Becker, Amy Deng, Katharyn Garcia, Max Hasin · Mar 18, 2025

Citations: 0
Expert Verification Automatic Metrics Tool Use General
  • Despite rapid progress on AI benchmarks, the real-world meaning of benchmark performance remains unclear.
  • To quantify the capabilities of AI systems in terms of human capabilities, we propose a new metric: 50%-task-completion time horizon.
A Survey on the Optimization of Large Language Model-based Agents

Shangheng Du, Jiabao Zhao, Jinxin Shi, Zhentao Xie, Xin Jiang, Yanhong Bai · Mar 16, 2025

Citations: 0
Simulation Env Long Horizon General
  • With the rapid development of Large Language Models (LLMs), LLM-based agents have been widely adopted in various fields, becoming essential for autonomous decision-making and interactive tasks.
  • However, current work typically relies on prompt design or fine-tuning strategies applied to vanilla LLMs, which often leads to limited effectiveness or suboptimal performance in complex agent-related environments.
Steering Dialogue Dynamics for Robustness against Multi-turn Jailbreaking Attacks

Hanjiang Hu, Alexander Robey, Changliu Liu · Feb 28, 2025

Citations: 0
Red Team Automatic Metrics General
  • To address this challenge, we propose a safety steering framework grounded in safe control theory, ensuring invariant safety in multi-turn dialogues.
  • Our method achieves invariant safety at each turn of dialogue by learning a safety predictor that accounts for adversarial queries, preventing potential context drift toward jailbreaks.
Can Multimodal LLMs Perform Time Series Anomaly Detection?

Xiongxiao Xu, Haoran Wang, Yueqing Liang, Philip S. Yu, Yue Zhao, Kai Shu · Feb 25, 2025

Citations: 0
Automatic Metrics Multi Agent Medicine
  • One natural way for humans to detect time series anomalies is through visualization and textual description.
  • To address the gap, we build a VisualTimeAnomaly benchmark to comprehensively investigate zero-shot capabilities of MLLMs for TSAD, progressively from point-, range-, to variate-wise anomalies, and extends to irregular sampling conditions.
Moving Beyond Medical Exams: A Clinician-Annotated Fairness Dataset of Real-World Tasks and Ambiguity in Mental Healthcare

Max Lamparth, Declan Grabb, Amy Franks, Scott Gershan, Kaitlyn N. Kunstman, Aaron Lulla · Feb 22, 2025

Citations: 0
Pairwise PreferenceExpert Verification Automatic Metrics MedicineCoding
  • Current medical language model (LM) benchmarks often over-simplify the complexities of day-to-day clinical practice tasks and instead rely on evaluating LMs on multiple-choice board exam questions.
  • This design enables systematic evaluations of model performance and bias by studying how demographic factors affect decision-making.
SEFL: A Framework for Generating Synthetic Educational Assignment Feedback with LLM Agents

Mike Zhang, Amalie Pernille Dilling, Léon Gondelman, Niels Erik Ruan Lyngdorf, Euan D. Lindsay, Johannes Bjerva · Feb 18, 2025

Citations: 0
Critique Edit Automatic Metrics General
  • Through comprehensive evaluations with three LLM judges and three human experts, across a subset of 900 outputs, we demonstrate that SEFL-tuned models outperform both their untuned counterparts and an existing baseline in terms of feedback
  • The potential for societal impact is reinforced by extensive qualitative comments and ratings from human stakeholders -- both students and higher education instructors.
Dialogue is Better Than Monologue: Instructing Medical LLMs via Strategical Conversations

Zijie Liu, Xinyu Zhao, Jie Peng, Zhuangdi Zhu, Qingyu Chen, Kaidi Xu · Jan 29, 2025

Citations: 0
Automatic MetricsSimulation Env Medicine
  • These tuning methods and benchmarks overlook critical aspects like evidence-based reasoning and handling distracting information.
  • To bridge this gap, we introduce a novel benchmark that simulates real-world diagnostic scenarios, integrating noise and difficulty levels aligned with USMLE standards.
Just KIDDIN: Knowledge Infusion and Distillation for Detection of INdecent Memes

Rahul Garg, Trilok Padhi, Hemang Jain, Ugur Kursuncu, Ponnurangam Kumaraguru · Nov 19, 2024

Citations: 0
Automatic MetricsSimulation Env General
  • Experimental results from our study on two hate speech benchmark datasets demonstrate superior performance over the state-of-the-art baselines across AU-ROC, F1, and Recall with improvements of 1.1%, 7%, and 35%, respectively.

Protocol Hubs

Expert Verification Papers (25) CS.CL + Expert Verification Papers (20) Pairwise Preference Papers (70) CS.CL + Pairwise Preference Papers (62) CS.AI + Expert Verification Papers (15) CS.AI + Pairwise Preference Papers (42) Rubric Rating Papers (17) CS.CL + Rubric Rating Papers (16) General + Pairwise Preference Papers (43) Expert Verification Or Rubric Rating Papers (39) CS.CL + Math Papers (84) Long Horizon Papers (82) CS.CL + Human Eval Papers (35) CS.CL + Long Horizon Papers (58) Expert Verification + Medicine Papers (11) Human Eval Papers (38)

Benchmark Hubs

Metric Hubs

Daily Archives