NeurIPS 2026 reviews are up: reviews, scores, rebuttals and meta-reviews for 7,877 papers. See the highest rated
acceptodds: the papers of ICLR 2027
2,182 results for “keyword:"inference"” in ICLR 2027. 456 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]Language Models as Information Ratchets: An Information-Thermodynamic Perspective on In-Context Learning▲3.0k2.33
[PDF]EvoComp-2: Learning to Compress Audio-Video Tokens for Omnimodal Large Language Models via Topic-Guided Evolutionary Labeling▼4.3k2.00
[PDF]Beyond Superficial Cues: Mitigating Popularity-Amplified Presentation Bias in Multimodal Recommendation▲3.0k2.33
[PDF]**StatOrch: Evidence-Conditioned Stagewise Inference Under Unknown Distribution Shift**▼1361.33
[PDF]Attend at the Right Resolution: Head-Adaptive Sparse Attention via Query–Key Interaction Structure▲3.0k2.33
[PDF]Support Selection Beyond Smooth DAG Exactness: Completion Geometry, Score Margins, and Selective Certificates▼5.8k2.33
[PDF]HeteroLoRA-MoE: Heterogeneous Low-Rank Experts with Diverse Projection Patterns and Efficient GPU Kernels▼4.3k2.00
[PDF]MedShard: Task-Structured Inference Improves Clinical Quality Control across Model Scales▼1611.50
[PDF]Compression-Aware Fine-Tuning: Adapting Language Model Weights to a Frozen Cache Bottleneck▼4.3k2.00
[PDF]ACT-SNN: Activity-Constrained Two-Stage Conversion for Efficient Spiking Large Language Models▲3922.67
[PDF]Gradient Descent on Tokens Is Sparse: Understanding the Computational Structure of Differentiable Token Optimization▼4.3k2.00
[PDF]When Personalization Compromises Truth: Joint Control of User Alignment and Factuality in Large Language Models▲3.0k2.33
Withdrawn or desk rejected (456)
[PDF]CCQ: Compensation-Conditioned Quadratic Selection for Absorbing Skipped Experts in MoE LLMsWithdrawn2.00
[PDF]TFAG-OVD: Training-Free Attribute-Guided Open-Vocabulary Detection via LLM Semantic CompositionWithdrawn1.67
[PDF]Guided Reasoning Across ISA Dialects: Inference-Time Control for GPU Assembly TranslationWithdrawn1.67
[PDF]TYPED AMPLITUDE COMPOSITION FOR REUSABLE INFERENCE: SEMANTICS, ERROR BOUNDS, AND EXACT CORRECTIONWithdrawn
[PDF]Graphs are Evidence, Not CircuitsDesk rejected