---

1. Multimodal Unlearning Across Vision, Language, Video, and Audio: Survey of Methods, Datasets, and Benchmarks
   Annual Meeting of the Association for Computational Linguistics2026-07-08alphaXiv arXiv S2
2. PACE: A Proxy for Agentic Capability Evaluation
   2026-07-02alphaXiv arXiv S2
3. Decomposer: Learning to Decompile Symbolic Music to Programs
   2026-07-02alphaXiv arXiv S2
4. PPT-Eval: A Benchmark for Computer-Use Agents on PowerPoint Tasks
   arXiv.org2026-06-30alphaXiv arXiv S2
5. Discretizing Reward Models
   arXiv.org2026-06-19alphaXiv arXiv S2
6. Enhancing Software Engineering Through Closed-Loop Memory Optimization
   arXiv.org2026-06-04alphaXiv arXiv S2
7. On the limits and opportunities of AI reviewers: Reviewing the reviews of Nature-family papers with 45 expert scientists
   arXiv.org2026-05-20alphaXiv arXiv S2
8. Reinforcing Human Behavior Simulation via Verbal Feedback
   arXiv.org2026-05-19alphaXiv arXiv S2
9. Soohak: A Mathematician-Curated Benchmark for Evaluating Research-level Math Capabilities of LLMs
   arXiv.org2026-05-09alphaXiv arXiv S2
10. Recursive Agent Optimization
    arXiv.org2026-05-07alphaXiv arXiv S2
11. Asking What Matters: Reward-Driven Clarification for Software Engineering Tasks
    arXiv.org2026-04-16alphaXiv arXiv S2
12. What do Language Models Learn and When? The Implicit Curriculum Hypothesis
    arXiv.org2026-04-09alphaXiv arXiv S2
13. Gym-Anything: Turn any Software into an Agent Environment
    arXiv.org2026-04-07alphaXiv arXiv S2
14. IDIOLEX: Unified and Continuous Representations for Idiolectal and Stylistic Variation
    arXiv.org2026-04-06alphaXiv arXiv S2
15. Effective Strategies for Asynchronous Software Engineering Agents
    arXiv.org2026-03-23alphaXiv arXiv S2
16. Reasoning over mathematical objects: on-policy reward modeling and test time aggregation
    arXiv.org2026-03-19alphaXiv arXiv S2
17. CodeScout: An Effective Recipe for Reinforcement Learning of Code Search Agents
    arXiv.org2026-03-18alphaXiv arXiv S2
18. CUBE: A Standard for Unifying Agent Benchmarks
    arXiv.org2026-03-16alphaXiv arXiv S2
19. Mind the Sim2Real Gap in User Simulation for Agentic Tasks
    arXiv.org2026-03-11alphaXiv arXiv S2
20. A Rubric-Supervised Critic from Sparse Real-World Outcomes
    arXiv.org2026-03-04alphaXiv arXiv S2
21. Real-Time Generation of Game Video Commentary with Multimodal LLMs: Pause-Aware Decoding Approaches
    arXiv.org2026-03-03alphaXiv arXiv S2
22. How Well Does Agent Development Reflect Real-World Work?
    arXiv.org2026-03-01alphaXiv arXiv S2
23. Modeling Distinct Human Interaction in Web Agents
    arXiv.org2026-02-19alphaXiv arXiv S2
24. Hybrid-Gym: Training Coding Agents to Generalize Across Tasks
    arXiv.org2026-02-18alphaXiv arXiv S2
25. Gained in Translation: Privileged Pairwise Judges Enhance Multilingual Reasoning
    Annual Meeting of the Association for Computational Linguistics2026-01-26alphaXiv arXiv S2
26. Massively Multilingual Joint Segmentation and Glossing
    Annual Meeting of the Association for Computational Linguistics2026-01-16alphaXiv arXiv S2
27. CAIRE: Cultural Attribution of Images with Retrieval
    Conference of the European Chapter of the Association for Computational Linguistics2026S2
28. Training Versatile Coding Agents in Synthetic Environments
    arXiv.org2025-12-13alphaXiv arXiv S2
29. On the Interplay of Pre-Training, Mid-Training, and RL on Reasoning Language Models
    arXiv.org2025-12-08alphaXiv arXiv S2
30. ClusterFusion: Hybrid Clustering with Embedding Guidance and LLM Adaptation
    arXiv.org2025-12-04alphaXiv arXiv S2
31. RefineBench: Evaluating Refinement Capability of Language Models via Checklists
    arXiv.org2025-11-27alphaXiv arXiv S2
32. Unsupervised Discovery of Long-Term Spatiotemporal Periodic Workflows in Human Activities
    IEEE Workshop/Winter Conference on Applications of Computer Vision2025-11-18alphaXiv arXiv S2
33. EDIT-Bench: Evaluating LLM Abilities to Perform Real-World Instructed Code Edits
    arXiv.org2025-11-06alphaXiv arXiv S2
34. The OpenHands Software Agent SDK: A Composable and Extensible Foundation for Production Agents
    arXiv.org2025-11-05alphaXiv arXiv S2
35. Training Proactive and Personalized LLM Agents
    arXiv.org2025-11-04alphaXiv arXiv S2
36. Oolong: Evaluating Long Context Reasoning and Aggregation Capabilities
    arXiv.org2025-11-04alphaXiv arXiv S2
37. Accumulating Context Changes the Beliefs of Language Models
    arXiv.org2025-11-03alphaXiv arXiv S2
38. The Tool Decathlon: Benchmarking Language Agents for Diverse, Realistic, and Long-Horizon Task Execution
    arXiv.org2025-10-29alphaXiv arXiv S2
39. Agent Data Protocol: Unifying Datasets for Diverse, Effective Fine-tuning of LLM Agents
    arXiv.org2025-10-28alphaXiv arXiv S2
40. How Do AI Agents Do Human Work? Comparing AI and Human Workflows Across Diverse Occupations
    arXiv.org2025-10-26alphaXiv arXiv S2
41. TOM-SWE: User Mental Modeling For Software Engineering Agents
    arXiv.org2025-10-24alphaXiv arXiv S2
42. TowerVision: Understanding and Improving Multilinguality in Vision-Language Models
    arXiv.org2025-10-22alphaXiv arXiv S2
43. Prompt-MII: Meta-Learning Instruction Induction for LLMs
    arXiv.org2025-10-19alphaXiv arXiv S2
44. Midtraining Bridges Pretraining and Posttraining Distributions
    arXiv.org2025-10-16alphaXiv arXiv S2
45. MERLIN: A Testbed for Multilingual Multimodal Entity Recognition and Linking
    Transactions of the Association for Computational Linguistics2025-10-16alphaXiv arXiv S2
46. How can we assess human-agent interactions? Case studies in software agent design
    arXiv.org2025-10-10alphaXiv arXiv S2
47. Grounding Multilingual Multimodal LLMs With Cultural Knowledge
    Conference on Empirical Methods in Natural Language Processing2025-08-10alphaXiv arXiv S2
48. Devstral: Fine-tuning Language Models for Coding Agent Applications
    arXiv.org2025-08-08alphaXiv arXiv S2
49. Checklists Are Better Than Reward Models For Aligning Language Models
    Neural Information Processing Systems2025-07-24alphaXiv arXiv S2
50. Code with Me or for Me? How Increasing AI Automation Transforms Developer Workflows
    International Conference on Human Factors in Computing Systems2025-07-10alphaXiv arXiv S2
51. OpenAgentSafety: A Comprehensive Framework for Evaluating Real-World AI Agent Safety
    arXiv.org2025-07-08alphaXiv arXiv S2
52. Does Math Reasoning Improve General LLM Capabilities? Understanding Transferability of LLM Reasoning
    arXiv.org2025-07-01alphaXiv arXiv S2
53. ZINA: Multimodal Fine-grained Hallucination Detection and Editing
    arXiv.org2025-06-16alphaXiv arXiv S2
54. CAIRe: Cultural Attribution of Images by Retrieval-Augmented Evaluation
    arXiv.org2025-06-10alphaXiv arXiv S2
55. Go-Browse: Training Web Agents with Structured Exploration
    arXiv.org2025-06-04alphaXiv arXiv S2
56. Coding Agents with Multimodal Browsing are Generalist Problem Solvers
    Conference of the European Chapter of the Association for Computational Linguistics2025-06-03alphaXiv arXiv S2
57. BehaviorBox: Automated Discovery of Fine-Grained Performance Differences Between Language Models
    Annual Meeting of the Association for Computational Linguistics2025-06-02alphaXiv arXiv S2
58. FieldWorkArena: Agentic AI Benchmark for Real Field Work Tasks
    arXiv.org2025-05-26alphaXiv arXiv S2
59. The CoT Encyclopedia: Analyzing, Predicting, and Controlling how a Reasoning Model will Think
    arXiv.org2025-05-15alphaXiv arXiv S2
60. VisualPuzzles: Decoupling Multimodal Reasoning Evaluation from Domain Knowledge
    arXiv.org2025-04-14alphaXiv arXiv S2
61. Do LLMs Understand Your Translations? Evaluating Paragraph-level MT with Question Answering
    arXiv.org2025-04-10alphaXiv arXiv S2
62. SkillWeaver: Web Agents can Self-Improve by Discovering and Honing Skills
    arXiv.org2025-04-09alphaXiv arXiv S2
63. Inducing Programmatic Skills for Agentic Tasks
    arXiv.org2025-04-09alphaXiv arXiv S2
64. M-Prometheus: A Suite of Open Multilingual LLM Judges
    arXiv.org2025-04-07alphaXiv arXiv S2
65. Scaling Evaluation-Time Compute with Reasoning Models as Evaluators
    Annual Meeting of the Association for Computational Linguistics2025-03-25alphaXiv arXiv S2
66. Overtrained Language Models Are Harder to Fine-Tune
    International Conference on Machine Learning2025-03-24alphaXiv arXiv S2
67. Benchmarking Failures in Tool-Augmented Language Models
    North American Chapter of the Association for Computational Linguistics2025-03-18alphaXiv arXiv S2
68. Efficient Many-Shot In-Context Learning with Dynamic Block-Sparse Attention
    Annual Meeting of the Association for Computational Linguistics2025-03-11alphaXiv arXiv S2
69. Not-Just-Scaling Laws: Towards a Better Understanding of the Downstream Impact of Language Model Design Decisions
    Conference on Empirical Methods in Natural Language Processing2025-03-05alphaXiv arXiv S2
70. ESPnet-SpeechLM: An Open Speech Language Model Toolkit
    North American Chapter of the Association for Computational Linguistics2025-02-21alphaXiv arXiv S2
71. Ambig-SWE: Interactive Agents to Overcome Underspecificity in Software Engineering
    2025-02-18alphaXiv arXiv S2
72. The Danger of Overthinking: Examining the Reasoning-Action Dilemma in Agentic Tasks
    arXiv.org2025-02-12alphaXiv arXiv S2
73. Demystifying Long Chain-of-Thought Reasoning in LLMs
    arXiv.org2025-02-05alphaXiv arXiv S2
74. Multilingual Machine Translation with Open Large Language Models at Practical Scale: An Empirical Study
    North American Chapter of the Association for Computational Linguistics2025-02-04alphaXiv arXiv S2
75. CowPilot: A Framework for Autonomous and Human-Agent Collaborative Web Navigation
    North American Chapter of the Association for Computational Linguistics2025-01-28alphaXiv arXiv S2
76. AutoPresent: Designing Structured Visuals from Scratch
    Computer Vision and Pattern Recognition2025-01-01alphaXiv arXiv S2
77. Evaluating Numeracy of Language Models as a Natural Language Inference Task
    North American Chapter of the Association for Computational Linguistics2025S2
78. AgentDiagnose: An Open Toolkit for Diagnosing LLM Agent Trajectories
    Conference on Empirical Methods in Natural Language Processing2025S2
79. Demystifying Long Chain-of-Thought Reasoning
    International Conference on Machine Learning2025S2
80. MAmmoTH-VL: Eliciting Multimodal Reasoning with Instruction Tuning at Scale
    Annual Meeting of the Association for Computational Linguistics2025S2
81. Pangea: A Fully Open Multilingual Multimodal LLM for 39 Languages
    International Conference on Learning Representations2025S2
82. Synthetic Data in the Era of Large Language Models
    Annual Meeting of the Association for Computational Linguistics2025S2
83. Scaling Evaluation-time Compute with Reasoning Models as Process Evaluators
    arXiv.org2025S2
84. Training Software Engineering Agents and Verifiers with SWE-Gym
    International Conference on Machine Learning2024-12-30alphaXiv arXiv S2
85. TheAgentCompany: Benchmarking LLM Agents on Consequential Real World Tasks
    Advances in Neural Information Processing Systems 382024-12-18alphaXiv arXiv S2
86. Towards Automatic Evaluation for Image Transcreation
    North American Chapter of the Association for Computational Linguistics2024-12-18alphaXiv arXiv S2
87. HILITE: Human-in-the-loop Interactive Tool for Image Editing
    BigData Congress [Services Society]2024-12-15S2
88. MAmmoTH-VL: Eliciting Multimodal Reasoning with Instruction Tuning at Scale
    arXiv.org2024-12-06alphaXiv arXiv S2
89. The BrowserGym Ecosystem for Web Agent Research
    Trans. Mach. Learn. Res.2024-12-06alphaXiv arXiv S2
90. Evaluating Language Models as Synthetic Data Generators
    Annual Meeting of the Association for Computational Linguistics2024-12-04alphaXiv arXiv S2
91. OpenScholar: Synthesizing Scientific Literature with Retrieval-augmented LMs
    arXiv.org2024-11-21alphaXiv arXiv S2
92. What Goes Into a LM Acceptability Judgment? Rethinking the Impact of Frequency and Length
    North American Chapter of the Association for Computational Linguistics2024-11-04alphaXiv arXiv S2
93. JMMMU: A Japanese Massive Multi-discipline Multimodal Understanding Benchmark for Culture-aware Evaluation
    North American Chapter of the Association for Computational Linguistics2024-10-22alphaXiv arXiv S2
94. Pangea: A Fully Open Multilingual Multimodal LLM for 39 Languages
    arXiv.org2024-10-21alphaXiv arXiv S2
95. Beyond Browsing: API-Based Web Agents
    Annual Meeting of the Association for Computational Linguistics2024-10-21alphaXiv arXiv S2
96. NaturalBench: Evaluating Vision-Language Models on Natural Adversarial Samples
    Neural Information Processing Systems2024-10-18alphaXiv arXiv S2
97. Harnessing Webpage UIs for Text-Rich Visual Understanding
    International Conference on Learning Representations2024-10-17alphaXiv arXiv S2
98. Stereotype or Personalization? User Identity Biases Chatbot Recommendations
    Annual Meeting of the Association for Computational Linguistics2024-10-08alphaXiv arXiv S2
99. Better Instruction-Following Through Minimum Bayes Risk
    International Conference on Learning Representations2024-10-03alphaXiv arXiv S2
100. Synatra: Turning Indirect Knowledge into Direct Demonstrations for Digital Agents at Scale
     Neural Information Processing Systems2024-09-24alphaXiv arXiv S2
101. Agent Workflow Memory
     International Conference on Machine Learning2024-09-11alphaXiv arXiv S2
102. MMMU-Pro: A More Robust Multi-discipline Multimodal Understanding Benchmark
     Annual Meeting of the Association for Computational Linguistics2024-09-04alphaXiv arXiv S2
103. Program-Aided Reasoners (Better) Know What They Know
     North American Chapter of the Association for Computational Linguistics2024S2
104. VisualWebArena: Evaluating Multimodal Agents on Realistic Visual Web Tasks
     Annual Meeting of the Association for Computational Linguistics2024S2
105. An image speaks a thousand words, but can everyone listen? On translating images for cultural relevance
     arXiv.org2024S2
106. Alignment for Honesty
     Advances in Neural Information Processing Systems 372024S2
107. An Incomplete Loop: Deductive, Inductive, and Abductive Learning in Large Language Models
     arXiv.org2024S2
108. OpenDevin: An Open Platform for AI Software Developers as Generalist Agents
     arXiv.org2024S2
109. RAGGED: Towards Informed Design of Retrieval Augmented Generation Systems
     arXiv.org2024S2
110. CMU’s IWSLT 2024 Offline Speech Translation System: A Cascaded Approach For Long-Form Robustness
     International Workshop on Spoken Language Translation2024S2
111. Divergences between Language Models and Human Brains
     Neural Information Processing Systems2024S2
112. GlossLM: Multilingual Pretraining for Low-Resource Interlinear Glossing
     arXiv.org2024S2
113. DiffusER: Diffusion via Edit-based Reconstruction
     International Conference on Learning Representations2023S2
114. SigMoreFun Submission to the SIGMORPHON Shared Task on Interlinear Glossing
     Special Interest Group on Computational Morphology and Phonology Workshop2023S2
115. Recent Developments in Computational Typology and Multilingual Natural Language Processing
     S2
116. Towards Scalable Oversight: Meta-Evaluation of LLMs as Evaluators via Agent Debate
     S2
117. Dynamic Summarization Length Control via Model Merging
     S2
118. A Survey of Cross-Lingual Alignment: Definitions, Methods, Future
     S2
119. CoderGen: Towards Domain-Specific Code Generation of Large Language Models
     S2
120. GenAI-Bench: A Holistic Benchmark for Compositional Text-to-Visual Generation
     S2
121. P OSITION : A GENTIC S YSTEMS S HOULD BE G ENERAL
     S2
122. Position: Humans are Missing from AI Coding Agent Research
     S2
123. Cracks in the Foundation: Seemingly Minor Architectural Choices Impact Long Context Extension
     S2
124. Multi-Task Learning with Self-Supervised Objectives can Improve Worst-Group Outcomes
     S2