Image de couverture de livre

Don't Waste Mistakes: Leveraging Negative RL-Groups via Confidence Reweighting

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Feng, Yunzhen, Jain, Parag, Hartshorn, Anthony, Duan, Yaqi, Kempe, Julia
Format:	Preprint
Publié:	2025
Sujets:	Machine Learning
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

What Characterizes Effective Reasoning? Revisiting Length, Review, and Structure of CoT
par: Feng, Yunzhen, et autres
Publié: (2025)

PILAF: Optimal Human Preference Sampling for Reward Modeling
par: Feng, Yunzhen, et autres
Publié: (2025)

Model Collapse Demystified: The Case of Regression
par: Dohmatob, Elvis, et autres
Publié: (2024)

Strong Model Collapse
par: Dohmatob, Elvis, et autres
Publié: (2024)

Beyond Model Collapse: Scaling Up with Synthesized Data Requires Verification
par: Feng, Yunzhen, et autres
Publié: (2024)

A Tale of Tails: Model Collapse as a Change of Scaling Laws
par: Dohmatob, Elvis, et autres
Publié: (2024)

Attacking Bayes: On the Adversarial Robustness of Bayesian Neural Networks
par: Feng, Yunzhen, et autres
Publié: (2024)

Don't Waste It: Guiding Generative Recommenders with Structured Human Priors via Multi-Head Decoding
par: Zhang, Yunkai, et autres
Publié: (2025)

Don't Trade Off Safety: Diffusion Regularization for Constrained Offline RL
par: Guo, Junyu, et autres
Publié: (2025)

Don't Waste Your Time: Early Stopping Cross-Validation
par: Bergman, Edward, et autres
Publié: (2024)

Efficient RL Training for LLMs with Experience Replay
par: Arnal, Charles, et autres
Publié: (2026)

Trust, or Don't Predict: Introducing the CWSA Family for Confidence-Aware Model Evaluation
par: Shahnazari, Kourosh, et autres
Publié: (2025)

Don't Be So Positive: Negative Step Sizes in Second-Order Methods
par: Shea, Betty, et autres
Publié: (2024)

Knowing What You Know Is Not Enough: Large Language Model Confidences Don't Align With Their Actions
par: Pal, Arka, et autres
Publié: (2025)

Experts Don't Cheat: Learning What You Don't Know By Predicting Pairs
par: Johnson, Daniel D., et autres
Publié: (2024)

Leveraging Sparsity for Sample-Efficient Preference Learning: A Theoretical Perspective
par: Yao, Yunzhen, et autres
Publié: (2025)

Angles Don't Lie: Unlocking Training-Efficient RL Through the Model's Own Signals
par: Wang, Qinsi, et autres
Publié: (2025)

Don't flatten, tokenize! Unlocking the key to SoftMoE's efficacy in deep RL
par: Sokar, Ghada, et autres
Publié: (2024)

Dual-Stage Reweighted MoE for Long-Tailed Egocentric Mistake Detection
par: Han, Boyu, et autres
Publié: (2025)

Don't Fool Me Twice: Adapting to Adversity in the Wild with Experience-Driven Reasoning
par: Ravie, Navin Sriram, et autres
Publié: (2026)

If You Don't Understand It, Don't Use It: Eliminating Trojans with Filters Between Layers
par: Hernandez, Adriano
Publié: (2024)

CurveRL: Principled Distribution-Aware Context Reweighting for LLM Reasoning
par: Sun, Ke, et autres
Publié: (2026)

Don't Freeze, Don't Crash: Extending the Safe Operating Range of Neural Navigation in Dense Crowds
par: Zhang, Jiefu, et autres
Publié: (2026)

Emergent properties with repeated examples
par: Charton, François, et autres
Publié: (2024)

When Models Don't Collapse: On the Consistency of Iterative MLE
par: Barzilai, Daniel, et autres
Publié: (2025)

Position: Don't be Afraid of Over-Smoothing And Over-Squashing
par: Kormann, Niklas, et autres
Publié: (2026)

Grow, Don't Overwrite: Fine-tuning Without Forgetting
par: Adila, Dyah, et autres
Publié: (2026)

Don't Be Greedy, Just Relax! Pruning LLMs via Frank-Wolfe
par: Roux, Christophe, et autres
Publié: (2025)

Don't Pay Attention, PLANT It: Pretraining Attention via Learning-to-Rank
par: Roy, Debjyoti Saha, et autres
Publié: (2024)

Don't Stop Me Yet: Sampling Loss Minima via Dissipative Riemannian Mechanics
par: Jacobsen, Albert Kjøller, et autres
Publié: (2026)

Class Confidence Aware Reweighting for Long Tailed Learning
par: Jagati, Brainard Philemon, et autres
Publié: (2026)

We Still Don't Understand High-Dimensional Bayesian Optimization
par: Doumont, Colin, et autres
Publié: (2025)

Don't Stop Me Now: Embedding Based Scheduling for LLMs
par: Shahout, Rana, et autres
Publié: (2024)

Don't Forget Imagination!
par: Vityaev, Evgenii E., et autres
Publié: (2025)

Transformers Don't In-Context Learn Least Squares Regression
par: Hill, Joshua, et autres
Publié: (2025)

Don't Walk the Line: Boundary Guidance for Filtered Generation
par: Ball, Sarah, et autres
Publié: (2025)

Don't Explain Noise: Robust Counterfactuals for Randomized Ensembles
par: Forel, Alexandre, et autres
Publié: (2022)

ReducedLUT: Table Decomposition with "Don't Care" Conditions
par: Cassidy, Oliver, et autres
Publié: (2024)

Position: Measure Dataset Diversity, Don't Just Claim It
par: Zhao, Dora, et autres
Publié: (2024)

Don't Forget the Nonlinearity: Unlocking Activation Functions in Efficient Fine-Tuning
par: Yin, Bo, et autres
Publié: (2025)