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
760 results for “keyword:"post training"” in ICLR 2027. 135 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]Decoupling Capabilities via Sample-Level Geometric Routing for Parameter-Efficient Low-Resource Adaptation▲6.1k2.00
[PDF]CovMOLAgent: Execution-Grounded Policy Learning for Multi-Objective Covalent Molecule Optimization▼1.2k1.67
[PDF]SACQ: Structure-Aware Codebook Quantization for Ultra-Low-Bit Vision-Language-Action Models▼4.5k2.67
[PDF]Don’t Rewrite, But Correct: Closed-Loop Manifold Correction via RL Post-Training for Generative Driving Policies▲3.0k2.33
[PDF]GraphPTQ: Post-Training Quantization for Graph Models, Memory Savings and Generalization▼5.8k2.33
[PDF]The Alignment Paradox: How Post-Training Amplifies Confident Hallucinations in Language Models▼4.3k2.00
[PDF]Learn What’s Left, Not What’s Mastered: Saturation Aware Advantage Reweighting for Multi-Reward Policy Optimization▼5.8k2.33
[PDF]Visitation Is Not Supervision: D-Optimality-Guided Allocation for On-Policy Distillation▼5.8k2.33
[PDF]Recurrent Self-Improvement: Dynamic Cross-Loop On-Policy Distillation for Looped Language Models▼4.5k2.67
[PDF]What Survives Continued Training? A Longitudinal Benchmark for Emergent-Misalignment Defenses▼5.8k2.33
[PDF]ProbePO: Sparse Probes, Redistributed Credit for Critic-Free LLM Reinforcement Learning▼4.5k2.67
[PDF]From Post-Training Plasticity to Function-Targeted Adaptation in Large Language Models▼4.3k2.00
[PDF]Beyond Policy Rollouts: Embedding-Space Directional Exploration with Anchored Noise for LLM RLVR▼4.5k2.67
[PDF]Learning without Overwriting: A Theory of Self-Distillation and Supervised Fine-Tuning in Continual Reasoning▲6.8k3.67
[PDF]TRIM: Temporal-subspace Reconstruction Induced by Moments for Post-Training Structured Pruning of Spiking Neural Networks▲5.4k3.00
Withdrawn or desk rejected (135)
[PDF]Everything in Moderation: Interior Per-Domain Coverage Optima and Gap-Preserving Fixed-Budget Alignment in Multi-Domain Mid-TrainingDesk rejected
[PDF]Why Benign Post-Training Induces Safety Gate Collapse: Changes in Harmful Activation Geometry Relative to the Inherited Refusal GateWithdrawn1.67
[PDF]Hierarchical Safety-Oriented Dense RL with Refined Credit Assignment for Tokenized Multi-Agent Traffic SimulationWithdrawn1.75
[PDF]Decoupling KL and Trajectories: A Unified Perspective for SFT, DAgger, Offline RL, and OPD in LLM DistillationDesk rejected1.50