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
354 results for “keyword:"pretraining"” in ICLR 2027. 64 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]AudioWeb-365M: A Web-Scale Corpus and Scaling Studies for Contrastive Language-Audio Pretraining▲5103.00
[PDF]PyroAdapt: Adapting Wildfire Prediction under Spatial Heterogeneity and Temporal Shift▲3.0k2.33
[PDF]GRAFT-MOE: GROUP-GUIDED ROUTING ADAPTA- TION AND FUNCTIONAL TRANSFER FOR PRETRAINED MOE EXPANSION▼5.8k2.33
[PDF]Replay on Demand: An Emergent Curriculum for Balancing Adaptation and Forgetting in Continued Pretraining▼5.8k2.33
[PDF]AttentionInfluence: Training-Free Weak-to-Strong Pretraining Data Selection via Attention Head Masking▼4.5k2.67
[PDF]Quality Is Not One Thing: Auditing, Distilling, and Blending LLM-Judge Annotations for Pre-Training Data Selection▼1361.33
[PDF]Token Mass, Not Geometry: A Confound in Embedding-Based Data Selection for Language Model Pretraining▼1.3k1.75
[PDF]One Marginal Is Enough: What Doubly Stochastic Routing Buys Hyper-Connections, and What It Costs▲3.0k2.33
Withdrawn or desk rejected (64)
[PDF]The Variance Brain Foundation Models Forgot: A Third-Order Basis for Predicting Individual DifferencesWithdrawn1.67
[PDF]Learning to Be Fooled: pretraining grows knowledge but keeps the scar of framing susceptibilityDesk rejected
[PDF]Input Representations Meet Pretraining Objectives: A Controlled Study of Generative EHR Foundation ModelsWithdrawn1.67
[PDF]Scene-Graph Molmo: Adding Structure to Pretraining Yields Strong Vision-Language ModelsWithdrawn1.75