Publications

(2025). System-1.5 Reasoning: Traversal in Language and Latent Spaces with Dynamic Shortcuts. Advances in Neural Information Processing Systems (NeurIPS 2025).
(2025). WebMMU: A Benchmark for Multimodal Multilingual Website Understanding and Code Generation. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP 2025).
(2025). VCR: Pixel-Level Complex Reasoning by Restoring Occluded Text. The Thirteenth International Conference on Learning Representations (ICLR 2025).
(2025). STRICT: Stress-Test of Rendering Image Containing Text. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP 2025).
(2025). Rethinking Decentralized Learning: Towards More Realistic Evaluations with a Metadata-Agnostic Approach. ICLR 2025 Workshop on Modularity for Collaborative, Decentralized, and Continual Deep Learning.
(2025). R³Mem: Bridging Memory Retention and Retrieval via Reversible Compression. Findings of the Association for Computational Linguistics: ACL 2025.
(2025). R(^mbox3)Mem: Bridging Memory Retention and Retrieval via Reversible Compression. CoRR.
(2025). MAP: Low-compute Model Merging with Amortized Pareto Fronts via Quadratic Approximation.
(2025). LongRecipe: Recipe for Efficient Long Context Generalization in Large Language Models. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025).
(2025). GraphOmni: A Comprehensive and Extendable Benchmark Framework for Large Language Models on Graph-theoretic Tasks. arXiv preprint arXiv: 2504.12764.