Research
A paper is listed when one of its authors is affiliated with the lab. Data: OpenAlex.
MAViQ: Memory-Augmented Visual Querying for Multi-Turn MLLM JailbreaksAlibaba · Conference on Empirical Methods in Natural Language Processing (EMNLP)Jan 1Latent Visual Sensitive Fingerprinting for Black-Box Vision–Language Model Tamper DetectionMicrosoft · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)Jan 1FormalImG: Evaluating Structural Compositional Generalization for T2I ModelsAlibaba · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)Jan 1OmniLayout: A Schematic-Coupled Multimodal Benchmark for Geometric Reasoning in PCB LayoutMicrosoft · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)Jan 1Lost in Embedding: Recovering Information Loss in Dense RetrievalMicrosoft · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)Jan 1PerturbPrune: Guiding Sparsity Allocation by Probing Layer Stability Along Principal DirectionsAlibaba · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)Jan 1KG-RLVR: Reinforcing LLM Reasoning with Knowledge Graph-Derived Process RewardsiFlytek · Conference on Empirical Methods in Natural Language Processing (EMNLP)Jan 1TTML: Test-Time Memory Learning for Label-Free RLVRAlibaba · Conference on Empirical Methods in Natural Language Processing (EMNLP)Jan 1Speaking While Listening: A Survey and Empirical Audit of Full-Duplex Spoken Dialogue SystemsAlibaba · Conference on Empirical Methods in Natural Language Processing (EMNLP)Jan 1Deliberation on Demand: Selective Deep Thinking for Language AgentsMicrosoft · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)Jan 1From Decoding Knobs to Commit Frontiers: Understanding Trajectory Control in Discrete Diffusion Language ModelsAlibaba · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)Jan 1COMPASS: Critique-guided Optimization of MetaPolicies for Agentic Systems with Reinforcement LearningAlphabet · Workshop for Research on Agent Language Models (REALM)Jan 1Reasoning Confidence Is Not Policy-Invariant: Calibrating Prefix Success under Model and Budget ShiftAlibaba · Conference on Empirical Methods in Natural Language Processing (EMNLP)Jan 1Incorporating Explicit and Implicit Chain-of-Thought as Hybrid Reasoning Paradigm with Guided Latent Space ModelingAlibaba · Conference on Empirical Methods in Natural Language Processing (EMNLP)Jan 1TraceForge: A Human-Gated System for Auto-Hardening Tool-Using LLM Agents into Immutable Workflow DAGsMultiverse Computing · Conference on Empirical Methods in Natural Language Processing: System Demonstrations (EMNLP)Jan 1Improving Speech Generation via Multi-Neighbor Token Prediction for Neural Codec Language ModelsAlibaba · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)Jan 1Who Is Asking? Personalized and Tiered Memory for Long-Lived AI AssistantsMicrosoft · Conference on Empirical Methods in Natural Language Processing: System Demonstrations (EMNLP)Jan 1Process Rewarding via Solution-Space AlignmentMicrosoft · Conference on Empirical Methods in Natural Language Processing (EMNLP)Jan 1Beyond Query-Item Matching: Personalized Behavioral Relevance Modeling for E-commerce Search via Rubric-Guided Reinforcement LearningAlibaba · Conference on Empirical Methods in Natural Language Processing (EMNLP)Jan 1SocialArena: Evaluating Social Intelligence in Large Language ModelsAlibaba · Conference on Empirical Methods in Natural Language Processing (EMNLP)Jan 1GR2: Mitigating Granularity Collapse via Adaptive Query Routing in Generative RetrievalLG AI Research · Conference on Empirical Methods in Natural Language Processing (EMNLP)Jan 1Beyond Prompts: Discovering Jailbreak Motifs in Sparse Feature SpaceAlibaba · Conference on Empirical Methods in Natural Language Processing (EMNLP)Jan 1Simulating Democratic Deliberation: Representing Pluralistic Preferences through Electoral Systems in Multi-Agent LLMsGoogle DeepMind · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)Jan 1DSPP: Dataset-Specific Progressive Pruning for Efficient and Generalizable LLM DeploymentAlibaba · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)Jan 1Causal Abstention for Cost-Aware Language-Agent CollaborationAlibaba · Conference on Empirical Methods in Natural Language Processing (EMNLP)Jan 1