Research
A paper is listed when one of its authors is affiliated with the lab. Data: OpenAlex.
MT2KD: Toward a General-Purpose Encoder for Speech, Speaker, and Audio EventsAlphabet, Google DeepMind · IEEE Transactions on Audio Speech and Language Processing2 citationsJan 1, 2025Revisiting In-Context Learning with Long Context Language ModelsGoogle DeepMind · Findings of the Association for Computational Linguistics: ACL (Findings of ACL)2 citationsJan 1, 2025MDCure: A Scalable Pipeline for Multi-Document Instruction-FollowingAlphabet · Proceedings of the conference - Association for Computational Linguistics. Meeting2 citationsJan 1, 2025Magnet: Multi-turn Tool-use Data Synthesis and Distillation via Graph TranslationAlphabet · Proceedings of the conference - Association for Computational Linguistics. Meeting2 citationsJan 1, 2025Do Large Language Models Perform Latent Multi-Hop Reasoning without Exploiting Shortcuts?Alphabet · Findings of the Association for Computational Linguistics: ACL (Findings of ACL)2 citationsJan 1, 2025Reverse Thinking Makes LLMs Stronger ReasonersAlphabet, Google DeepMind · Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT)2 citationsJan 1, 2025Chatbot Arena Estimate: towards a generalized performance benchmark for LLM capabilitiesAlphabet · Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT)2 citationsJan 1, 2025An Efficient Rehearsal Scheme for Catastrophic Forgetting Mitigation during Multi-stage Fine-tuningAlphabet · Findings of the Association for Computational Linguistics: NAACL2 citationsJan 1, 2025Human-Data Interaction: Thinking Beyond Individual DatasetsAlphabet · interactions2 citationsJan 1, 2025Tracing Back the Malicious Clients in Poisoning Attacks to Federated LearningAlphabet · neural information processing systems2 citationsJan 1, 2025MoCha: Towards Movie-Grade Talking Character GenerationGoogle DeepMind, Meta · neural information processing systems1 citationJan 1, 2025PiKE: Adaptive Data Mixing for Large-Scale Multi-Task Learning Under Low Gradient ConflictsAlphabet · neural information processing systems1 citationJan 1, 2025Robust LLM Alignment via Distributionally Robust Direct Preference OptimizationAlphabet · neural information processing systems1 citationJan 1, 2025Matryoshka Pilot: Learning to Drive Black-Box LLMs with LLMsAlphabet, Google DeepMind · neural information processing systems1 citationJan 1, 2025Planning without Search: Refining Frontier LLMs with Offline Goal-Conditioned RLGoogle DeepMind · neural information processing systems1 citationJan 1, 2025The emergence of sparse attention: impact of data distribution and benefits of repetitionGoogle DeepMind · neural information processing systems1 citationJan 1, 2025EA3D: Online Open-World 3D Object Extraction from Streaming VideosAlphabet, Google DeepMind · neural information processing systems1 citationJan 1, 2025EvoLM: In Search of Lost Training Dynamics for Language Model ReasoningGoogle DeepMind · neural information processing systems1 citationJan 1, 2025Generative Trajectory Stitching through Diffusion CompositionGoogle DeepMind · neural information processing systems1 citationJan 1, 2025Tensor-Parallelism with Partially Synchronized ActivationsAlphabet · neural information processing systems1 citationJan 1, 2025A Implies B: Circuit Analysis in LLMs for Propositional Logical ReasoningAlphabet · neural information processing systems1 citationJan 1, 2025Vision Transformers with Self-Distilled RegistersAlphabet, Google DeepMind · neural information processing systems1 citationJan 1, 2025Exploring the limits of strong membership inference attacks on large language modelsAnthropic, Google DeepMind, OpenAI · neural information processing systems1 citationJan 1, 2025Skill-Driven Neurosymbolic State AbstractionsGoogle DeepMind · neural information processing systems1 citationJan 1, 2025DataRater: Meta-Learned Dataset CurationAlphabet, Google DeepMind · neural information processing systems1 citationJan 1, 2025