Research
A paper is listed when one of its authors is affiliated with the lab. Data: OpenAlex.
Dense SAE Latents Are Features, Not BugsGoogle DeepMind · neural information processing systems13 citationsJan 1, 2025Do LLMs Really Forget? Evaluating Unlearning with Knowledge Correlation and Confidence AwarenessAlphabet · neural information processing systems13 citationsJan 1, 2025Win Fast or Lose Slow: Balancing Speed and Accuracy in Latency-Sensitive Decisions of LLMsGoogle DeepMind, Microsoft, NVIDIA · neural information processing systems13 citationsJan 1, 2025RA-BLIP: Multimodal Adaptive Retrieval-Augmented Bootstrapping Language-Image Pre-TrainingAlibaba · IEEE Transactions on Multimedia13 citationsJan 1, 2025Qwen2.5-xCoder: Multi-Agent Collaboration for Multilingual Code Instruction TuningAlibaba · Proceedings of the conference - Association for Computational Linguistics. Meeting13 citationsJan 1, 2025TUMLU: A Unified and Native Language Understanding Benchmark for Turkic LanguagesMicrosoft · Proceedings of the conference - Association for Computational Linguistics. Meeting13 citationsJan 1, 2025ProgCo: Program Helps Self-Correction of Large Language ModelsAlibaba · Proceedings of the conference - Association for Computational Linguistics. Meeting13 citationsJan 1, 2025When2Call: When (not) to Call ToolsNVIDIA · Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT)13 citationsJan 1, 2025SeaExam and SeaBench: Benchmarking LLMs with Local Multilingual Questions in Southeast AsiaAlibaba · Findings of the Association for Computational Linguistics: NAACL13 citationsJan 1, 2025LibEvolutionEval: A Benchmark and Study for Version-Specific Code GenerationNVIDIA · Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT)13 citationsJan 1, 2025AIM 2024 Challenge on Efficient Video Super-Resolution for AV1 Compressed ContentMeta · Lecture notes in computer science13 citationsJan 1, 2025Escaping Collapse: The Strength of Weak Data for Large Language Model TrainingAlphabet · neural information processing systems13 citationsJan 1, 2025Scaling Embedding Layers in Language ModelsAlphabet · neural information processing systems13 citationsJan 1, 2025ICASSP 2024 Speech Signal Improvement ChallengeMicrosoft · IEEE Open Journal of Signal Processing13 citationsJan 1, 2025OpenGU: A Comprehensive Benchmark for Graph UnlearningAlibaba · neural information processing systems13 citationsJan 1, 2025Spectral Analysis of Diffusion Models with Application to Schedule DesignNVIDIA · neural information processing systems12 citationsJan 1, 2025RLZero: Direct Policy Inference from Language Without In-Domain SupervisionOpenAI · neural information processing systems12 citationsJan 1, 2025RePIC: Reinforced Post-Training for Personalizing Multi-Modal Language ModelsNVIDIA · neural information processing systems12 citationsJan 1, 2025Grounded Reinforcement Learning for Visual ReasoningAlphabet · neural information processing systems12 citationsJan 1, 2025GoT: Unleashing Reasoning Capability of MLLM for Visual Generation and EditingSenseTime · neural information processing systems12 citationsJan 1, 2025Towards Graph Foundation Models: Training on Knowledge Graphs Enables Transferability to General GraphsMicrosoft · neural information processing systems12 citationsJan 1, 2025EchoShot: Multi-Shot Portrait Video GenerationAlibaba · neural information processing systems12 citationsJan 1, 2025DGS-LRM: Real-Time Deformable 3D Gaussian Reconstruction From Monocular VideosAlphabet, Google DeepMind, Meta · neural information processing systems12 citationsJan 1, 2025The Emergence of Abstract Thought in Large Language Models Beyond Any LanguageAlibaba · neural information processing systems12 citationsJan 1, 2025Scaling Language-centric Omnimodal Representation LearningAlibaba · neural information processing systems12 citationsJan 1, 2025