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
64 results for “keyword:"mixed precision"” in ICLR 2027. 11 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]FROM ATTENTION SENSITIVITY TO LAYER ROLE: REVISITING MIXED-PRECISION QUANTIZATION OF TRANSFORMERS▼1.2k1.67
[PDF]Where Compression Hurts: Function-Space Sensitivity for Mixed-Precision Neural Network Quantization▼4.3k2.00
[PDF]KV4-RL: Unlocking Faster Asynchronous Rollouts for RLVR with INT4 KV Cache Quantization▲5103.00
[PDF]QGATE: Counterexample-Guided Precision Patching of Quantized LLMs with Fail-Closed Slice Reporting▲11.00
[PDF]LSMXQ: Layerwise Salience-Aware Mixed-Precision Quantization for Efficient LLM Inference▼1361.33
[PDF]QuMix: Query-Guided Mixed-Precision KV-Cache Quantization for Block Diffusion Language Models▲3.0k2.33
[PDF]Forward Similarity Does Not Guarantee Training Equivalence: Precision Placement in Differentiable Gram Solvers▼1.2k1.67
[PDF]Layerwise Error Attribution for Fast and Robust Mixed-Precision Post-Training Quantization▼5.8k2.33
[PDF]HiCAP-Q: Hierarchical Capability-Aware Mixed-Precision Quantization for Large Language Models▼1361.33
[PDF]BRINE: Matched-State Counterfactual Supervision for Risk-Aware Cognitive Routing in Underwater Robots▼1361.33
[PDF]Does Quantization Preserve How Language Models Answer? Causal-Equivariant Quantization for Symbol-Binding Circuits▼4.5k2.25
[PDF]Configuration-Valued Expert Importance for Mixed-Precision Quantization of Multimodal Mixture-of-Experts▼5.9k2.50
[PDF]Gradient Norms, Not Attention Mass: Cheap Proxies for Layer-Wise KV-Cache Bit Allocation▼1.2k1.67
[PDF]Off-Diagonal Hessian Sensitivity for Post-Training Quantization of Large Language Models▼4.3k2.00
[PDF]ThunderWorld: Efficient World Model Generation with Selective FP4 Refinement via Native GEMMs▲3.0k2.33
[PDF]SignRoundV2: Toward Closing the Performance Gap in Extremely Low-Bit Post-Training Quantization for LLMs▲3.0k2.33
[PDF]ShadowMoE: Graph-Conditioned Shadow Expert Representations for Mixed-Precision MoE Quantization▼1.2k1.67