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
[PDF]HEPH: 3-Bit Grid-Aligned Trellis Quantization of MXFP4 Mixture-of-Experts for Accurate, Hadamard-Free Serving▲20k2.67
[PDF]QAttrib-MoE: Runtime-Controlled Attribution of Cross-Precision Trajectory Divergence in Mixture-of-Experts Inference▼5.3k1.33
[PDF]AWSLQ: Activation-Wise Sparse-Ladder Quantization for Accurate W4A4 Single-Load Decoding▼4.3k2.00
[PDF]SCOPE: From Analytical Scale Oracles to Candidate Scale Search for NVFP4 Quantized Training▼4.5k2.67
[PDF]OSFP4: Joint Optimization of Diagonal Smoothing and Block Scales for NVFP4 Quantization▲6.1k2.00
[PDF]ScaleSweep: Accurate NVFP4 Post-Training Quantization of LLMs via Block Scale Initialization▼1.2k1.67
[PDF]LatentQuant: Preserving the Policy-Facing Latent Contract under NVFP4 VAE Quantization▼5.8k2.33
[PDF]ThunderWorld: Efficient World Model Generation with Selective FP4 Refinement via Native GEMMs▲3.0k2.33
[PDF]MXAttention: Data-Free Optimal Scaling and Pre-Normalization Quantization for MXFP4 Attention▼4.3k2.00