Research
A paper is listed when one of its authors is affiliated with the lab. Data: OpenAlex.
Prompt-Level Distillation: A Non-Parametric Alternative to Model Fine-Tuning for Efficient ReasoningAlphabet · Annual Meeting of the Association for Computational Linguistics (ACL)Jan 1Linear Probes Detect Task Format, Not Reasoning Mode in Language Model Hidden StatesAlphabet, Google DeepMind · Workshop on Trustworthy NLP (TrustNLP)Jan 1Understanding Emergent Misalignment via Feature Superposition GeometryGoogle DeepMind · Annual Meeting of the Association for Computational Linguistics (ACL)Jan 1IndicMMLU-Pro: Benchmarking Indic Large Language Models on Multi-Task Language UnderstandingAlphabet, Google DeepMind · Workshop on Generation, Evaluation and Metrics (GEM)Jan 1SEA-Guard: Culturally Grounded Multilingual Safeguard for Southeast AsiaAlphabet · Findings of the Association for Computational Linguistics: ACL (Findings of ACL)Jan 1Language Models Struggle to Use Representations Learned In-ContextAlphabet, Google DeepMind · Annual Meeting of the Association for Computational Linguistics (ACL)Jan 1FADE: Probing the Limits of VLMs on fine-grained OCRAlphabet · Workshop on Advances in Language and Vision Research (ALVR)Jan 1Self-Explaining Hate Speech Detection with Moral RationalesAlphabet · Findings of the Association for Computational Linguistics: ACL (Findings of ACL)Jan 1Speech Translation and Metrics in 2026: Findings of the IWSLT CampaignGoogle DeepMind · International Conference on Spoken Language Translation (IWSLT)Jan 1GraphMind: LLMs as Dynamic Knowledge Builders for Sequential Decision-MakingGoogle DeepMind · Findings of the Association for Computational Linguistics: ACL (Findings of ACL)Jan 1Location Not Found: Exposing Implicit Local and Global Biases in Multilingual LLMsAlphabet · Annual Meeting of the Association for Computational Linguistics (ACL)Jan 1MemoBench: Benchmarking World Modeling in Dynamically Changing EnvironmentsAlphabet · Lecture notes in computer scienceJan 1Detecting Safety Training Modification in Language Models via Activation AnalysisAlphabet · IEEE AccessJan 1Necro-reaper: Pruning away Dead Memory Traffic in Warehouse-Scale ComputersAlphabet · IEEE MicroJan 1Efficient Asynchronous Notification with Extended User Interrupts (xUI)Alphabet · IEEE MicroJan 1Fast and Accurate CPU Performance Modeling with Compositional Analytical–ML FusionAlphabet · IEEE MicroJan 1大規模言語モデルの現在地 ――世界の潮流と日本の動向――Google DeepMind · Denshi Jouhou Tsuushin Gakkai Tsuushin Sosaieti magajin/B-plus/Denshi Jouhou Tsuushin Gakkai Tsuushin Sosaeti magajinJan 1On the LSH Distortion of Ulam and Cayley SimilaritiesAlphabet · arXiv (Cornell University)Jan 1Problems of Scaling Large Language Model Inference During Simultaneous Operation of Multiple Autonomous AgentsAlphabet · International Journal of Advanced engineering Management and ScienceJan 1Federated Primitive-Preserving Audio Transformers for Non-Identifiable Infant Cry ClassificationAlphabet · IEEE AccessJan 1ELT: Elastic Looped Transformers for Visual GenerationGoogle DeepMind · Lecture notes in computer scienceJan 1Forecasting Motion in the WildGoogle DeepMind · Lecture notes in computer scienceJan 1Fisheye3R: Adapting Unified 3D Feed-Forward Foundation Models to Fisheye LensesAlphabet · Lecture notes in computer scienceJan 1The Emerging Paradigm of Geospatial Foundation Models: From Pre-training to Agentic ReasoningAlphabet · Lecture notes in computer scienceJan 1Multilingual Amnesia: On the Transferability of Unlearning in Multilingual LLMsAlphabet · Conference of the European Chapter of the Association for Computational LinguisticsJan 1