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
945 results for “keyword:"analysis"” in ICLR 2027. 246 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]SciReC: Diagnostic Evaluation of Multimodal Multi-Turn Relational Reasoning with Adaptive Interaction▼5.8k2.33
[PDF]Revisiting the 16× Pretraining Compute Gap between Diffusion and Autoregressive Language Models▼4.3k2.00
[PDF]One Capability or Many? Structural and Predictive Tests of Benchmark Validity Disagree About Economic Benchmarks for Frontier AI▼71.25
[PDF]Muon–Adam Training for Overparameterized Neural Networks: Convergence and Generalization▼4.3k2.00
[PDF]Not All Errors Are Equally Steerable: Predicting Error Types to Guide Activation Steering▼4.3k2.00
[PDF]Max-pooling Network Revisited: Analyzing the Role of Semantic Probability in Multiple Instance Learning for Hallucination Detection▲9.7k2.75
[PDF]GaitReader: From Large-Scale Healthy Gait Pretraining to Cross-Center Disease Recognition▲5.4k3.00
[PDF]Optimal Recursive Composition and Dyadic Phase Laws for Gradient Descent with Predetermined Stepsizes▲5103.00
[PDF]PharmEval: A Diagnostic Benchmark for Pharmacological Knowledge and Reasoning of Large Language Models▼5.8k2.33
[PDF]Normalization as Control: Batch Normalization Is the Deadbeat Corner of a Kalman-Tuned Normalizer Family▼4.3k2.00
[PDF]Fast and Quality-Preserving Non-prefix KV Cache Reuse for Large Language Model Serving▼5.8k2.33
[PDF]How Many Steps Does Flow Matching Really Need? A Frequency-Domain Analysis of Discretization Error in the Base-ODE Regime▼1.3k1.75
Withdrawn or desk rejected (246)
[PDF]ReLoC: Recursive Local Causal Modeling for Explaining and Predicting LLM-Based Multi-Agent BehaviorWithdrawn1.33
[PDF]CCQ: Compensation-Conditioned Quadratic Selection for Absorbing Skipped Experts in MoE LLMsWithdrawn2.00
[PDF]Probing Interaction Awareness in Trajectory Prediction with Counterfactual PerturbationsDesk rejected
[PDF]When Does Chain-of-Thought Hurt Spatial Answers? Length, Retrieval, and Content Controls in Vision-Language ModelsWithdrawn1.67