Carnegie Mellon University · 56 records

dateidtitleauthorsvenue↑
2026-11-132611.17834PaperCritic: Training AI Reviewers to Assess the Soundness of PapersHyun Ryu et al.arXiv32
2026-11-132611.16728EgoX: Egocentric Cross-Embodiment Manipulation with Embodiment DreamingYaru Niu et al.arXiv1
2026-11-092611.01362"PRISM" Is Not All You Need: Cross-Topic Name Collisions and Scientific Paper RetrievalJingjie Ning et al.arXiv6
2026-11-09W7147594723TokenDial: Continuous Attribute Control for Text-to-Video Generation in Visual Dial SpaceZhixuan Liu et al.ACM Transactions on Graphics
2026-10-312611.00883Random Is Almost Enough? Stable and Low-Cost Benchmark Compression via BestRandGeorge Ling et al.arXiv49
2026-10-30W7224843773BitterBuster: A virtual world for understanding naturalistic patch foragingJohn P. Kasarda et al.PLoS ONE
2026-10-292610.40564Training Foundation Models with a Single AlgorithmYuda Song et al.arXiv8
2026-10-282610.34714SearchHorizonBench: A Benchmark for Search Value in AI AgentsJingjie Ning et al.arXiv11
2026-10-282610.29003T-DIG: Text-Conditioned Diversity for Image GenerationSimran Khanuja et al.arXiv8
2026-10-282610.33245AdaPRA: Process Reward Agents with Adaptive Rubrics for Data ScienceAtharva Naik et al.arXiv3
2026-10-272610.34634AgentProofBench: Benchmarking Reusable Verifiers for Web AgentsJingjie Ning et al.arXiv12
2026-10-242610.33512Eat Your Precision: Precision Objective Mitigates Reward Hacking in Rubric-based RLHyun Ryu et al.arXiv47
2026-10-212610.23010What Makes a Good Symbol for Hybrid Discrete-Continuous Planning?Yiyuan Pan et al.arXiv2
2026-10-192610.19159Into the Unknown: Benchmarking AI agents on open scientific research problemsSeungone Kim et al.arXiv62
2026-10-142610.16811X-Ragent: Tool-Augmented Reinforcement Learning for Evidence-Grounded Chest X-Ray ReasoningYi Wang et al.arXiv34
2026-10-142610.18579Reading Right, Acting Wrong: Diagnosing Decision Failures of System One Models in Agent LoopsYubo Li et al.arXiv9
2026-10-082610.05034Agentic RAG Evaluation: Budget Allocation Across Questions, Trajectories, and ReadsJingjie Ning et al.arXiv10
2026-10-072610.01892Selection-Based Structured Reasoning: Toward Efficient Multimodal Search AgentsFeiyu Gavin Zhu et al.arXiv32
2026-10-022610.00314Predictive Credit: Measuring What Scientific Explanations Add to Experimental ForecastsJingjie Ning et al.arXiv11
2026-10-022610.00313Rules to Tools: Executable Checks for LLM Agents in Scientific ComputingJingjie Ning et al.arXiv5
2026-10-012609.32720BiasReducer: Adaptive Bias Mitigation for Reward ModelsShuang Liu et al.arXiv25
2026-10-012609.32600CUA-SWE: When Computer-Use Agents Meet Visual Software EngineeringPrince Zizhuang Wang et al.arXiv15
2026-09-292609.34313ControlScope: Workflow Revision and Reliability in LLM AgentsJingjie Ning et al.arXiv3
2026-09-282609.30222TrackEverything: Long Horizon Dense Tracking via De-Duplicating 3D Scene RepresentationsAyush Jain et al.arXiv14
2026-09-252609.28654Training Object Permanence in World ModelsHaotian Zhang et al.arXiv112
2026-09-242609.27490WhatWorkedBench: Benchmarking Experimental Understanding in AI AgentsJingjie Ning et al.arXiv15
2026-09-232609.26550JEV-as-a-Judge: Accept When Confident, Escalate When UnsureYubo Li et al.arXiv41
2026-09-162609.16204Decoy Direction Optimization: A Post-Hoc Defense Against LLM AbliterationAashiq Muhamed et al.arXiv4
2026-09-142609.11115Benchmark Radar: A Living Database and Search Engine for AI Benchmarks and EvaluationKoutian Wu et al.arXiv153
2026-09-102609.09219Scores Alone Do Not Prove Discovery: The Discovery Certification Protocol for Auditing AI Research AgentsJingjie Ning et al.arXiv13
2026-09-092609.08345CoVeR: Coverage-Based Token Pruning for Multi-View 3D Reasoning in VLMsNhat-Tan Bui et al.arXiv9
2026-09-092609.06966MOLE: Detecting Insider Threats in AI AgentsAashiq Muhamed et al.arXiv4
2026-08-252608.22856Same Agent, Different Answers: A Repeat-Aware Audit of Corpus-Induced Answer Churn in Retrieval-Augmented QAJingjie Ning et al.arXiv4
2026-08-192608.16977The Problem Is the Problem: Towards Scalable Mathematical DiscoveryZeyu Zheng et al.arXiv7
2026-08-172608.13760Amplified Does Not Mean Predictive: Reasoning Behaviors in Thinking ModelsJean de Dieu Nyandwi et al.arXiv4
2026-08-112608.03499WeClawArena: An Auditable Sandbox and Benchmark for Cross-User Agents Collaboration and Security in Human-Centered Agent NetworksPrince Zizhuang Wang et al.arXiv7
2026-07-272607.18142O-VAD: Industrial Video Anomaly Detection through Object-Centric Tracking and ReasoningMei Yuan et al.arXiv2
2026-06-302606.00097RocketSmith: Agentic Additive Manufacturing of High-Powered RocketsPeter Pak et al.arXiv6
2026-06-232606.23608Causal Discovery in the Era of AgentsYujia Zheng et al.arXiv8
2026-06-182606.16748MyPCBench: A Benchmark for Personally Intelligent Computer-Use AgentsLawrence Keunho Jang et al.arXiv6
2026-06-182606.09764iOSWorld: A Benchmark for Personally Intelligent Phone AgentsLawrence Keunho Jang et al.arXiv4
2026-06-092606.08960Hardening Agent Benchmarks with Adversarial Hacker-Fixer LoopsZiqian Zhong et al.arXiv2
2026-06-022606.02404K-BrowseComp: A Web Browsing Agent Benchmark Grounded in Korean ContextsNahyun Lee et al.arXiv56
2026-06-022605.29084Same Question, Different Source, Different Answer: Auditing Source-Dependence in Medical Multi-Source RAGYubo Li et al.arXiv1
2026-06-022605.29087The Chain Holds, the Answer Folds: Trace-Answer Dissociation in Reasoning Models Under Adversarial PressureYubo Li et al.arXiv1
2026-05-292605.24785PANDO: Efficient Multimodal AI Agents via Online Skill DistillationYubo Li et al.arXiv5
2026-05-292605.25220Multi-view Consistent 3D Gaussian Head Avatars 'without' Multi-view GenerationAviral Chharia et al.arXiv4
2026-05-292605.24786CONF-KV: Confidence-Aware KV Cache Eviction with Mixed-Precision Storage for Long-Horizon LLMYubo Li et al.arXiv3
2026-05-282605.26457Verus-SpecGym: An Agentic Environment for Evaluating Specification AutoformalizationAnmol Agarwal et al.arXiv4
2026-05-252605.17766LatentUMM: Dual Latent Alignment for Unified Multimodal ModelsYinyi Luo et al.arXiv7
2026-05-212605.20668On the limits and opportunities of AI reviewers: Reviewing the reviews of Nature-family papers with 45 expert scientistsSeungone Kim et al.arXiv13
2026-05-202605.19516Base Models Look Human To AI DetectorsYixuan Even Xu et al.arXiv1
2026-05-192605.15565AstraFlow: Dataflow-Oriented Reinforcement Learning for Agentic LLMsHaizhong Zheng et al.arXiv15
2026-05-192605.16865MixSD: Mixed Contextual Self-Distillation for Knowledge InjectionJiarui Liu et al.arXiv8
2026-05-142605.12733From Generalist to Specialist RepresentationYujia Zheng et al.arXiv1
2026-05-082605.05724Auto Research with Specialist Agents Develops Effective and Non-Trivial Training RecipesJingjie Ning et al.arXiv15