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
753 results for “keyword:"fine tuning"” in ICLR 2027. 158 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]Towards Mechanistically Understanding Why Memorized Knowledge Fails to Generalize in Large Language Model Finetuning▲6.1k2.00
[PDF]How Much Rank Does LoRA Need? Task-Dependent Rank-Error Bounds for Transformer Attention▼1.2k1.67
[PDF]Beyond the Current Observation: History-Informed Adaptation for Vision-Language-Action Models▼1361.33
[PDF]Beyond Non-Zero Initialization: Investigating Update Balance in Structured LoRA Initialization▼4.3k2.00
[PDF]TraceGen: Synthesizing Difficulty-Controllable Bugs for Training and Evaluating Code Repair Agents▼1361.33
[PDF]Statistical Benefits of Fine-Tuning from Pretrained Initialization in Diagonal Linear Networks▲3.0k2.33
[PDF]Benchmarking, Diagnosing, and Mitigating Temporal Positional Bias in Video Understanding▼5.8k2.33
[PDF]Don't Mask the Environment: Observation Supervision Changes How Agents Explore Under RL▲6.1k2.00
[PDF]Does Neighbour Relatedness Matter? Dissimilarity-Priority Packing Contradicts the Similarity Hypothesis in Supervised Fine-Tuning▼1361.33
[PDF]Enabling Downstream Adaptation of Spiking LLMs: From Zeroth-Order Learning to Cross-Resolution Transfer▼4.3k2.00
[PDF]Learning the Map, Not the Route: Curriculum, Capacity, and Transfer in Tool-Navigating Language Models▼4.5k2.67
[PDF]From Post-Training Plasticity to Function-Targeted Adaptation in Large Language Models▼4.3k2.00
[PDF]HQS-Tuning: A Hidden-State Quality-Guided Data Filtering Framework for Efficient Multimodal Instruction Fine-Tuning▼1.2k1.67
[PDF]Learning without Overwriting: A Theory of Self-Distillation and Supervised Fine-Tuning in Continual Reasoning▲6.8k3.67
[PDF]Quantized Evolution Strategies: High-precision Fine-tuning of Quantized LLMs at Low-precision Cost▲5103.00
[PDF]DoMA: Geometry-Aware Aggregation for Stable Continual Federated Fine-Tuning on the Heterogeneous Edge▼1.2k3.00
[PDF]Transform Coordinates Separate Vulnerability from Compensatory Capacity under Deployment Perturbations▼1.2k1.67
Withdrawn or desk rejected (158)
[PDF]When to Act, Not What to Act: Prospective Authorization for Tool-Using Language ModelsWithdrawn1.33
[PDF]Why Benign Post-Training Induces Safety Gate Collapse: Changes in Harmful Activation Geometry Relative to the Inherited Refusal GateWithdrawn1.67