Copertina

Cascade Reward Sampling for Efficient Decoding-Time Alignment

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Li, Bolian, Wang, Yifan, Lochab, Anamika, Grama, Ananth, Zhang, Ruqi
Natura:	Preprint
Pubblicazione:	2024
Soggetti:	Computation and Language Machine Learning
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Addressing Performance Saturation for LLM RL via Precise Entropy Curve Control
di: Li, Bolian, et al.
Pubblicazione: (2026)

Energy-Based Reward Models for Robust Language Model Alignment
di: Lochab, Anamika, et al.
Pubblicazione: (2025)

Uniform-Correct Policy Optimization: Breaking RLVR's Indifference to Diversity
di: Lochab, Anamika, et al.
Pubblicazione: (2026)

VERA-V: Variational Inference Framework for Jailbreaking Vision-Language Models
di: Liao, Qilin, et al.
Pubblicazione: (2025)

VERA: Variational Inference Framework for Jailbreaking Large Language Models
di: Lochab, Anamika, et al.
Pubblicazione: (2025)

ETA: Evaluating Then Aligning Safety of Vision Language Models at Inference Time
di: Ding, Yi, et al.
Pubblicazione: (2024)

Reward-Shifted Speculative Sampling Is An Efficient Test-Time Weak-to-Strong Aligner
di: Li, Bolian, et al.
Pubblicazione: (2025)

Entropy-MCMC: Sampling from Flat Basins with Ease
di: Li, Bolian, et al.
Pubblicazione: (2023)

Learning Self-Correction in Vision-Language Models via Rollout Augmentation
di: Ding, Yi, et al.
Pubblicazione: (2026)

DRIFT: Learning from Abundant User Dissatisfaction in Real-World Preference Learning
di: Wang, Yifan, et al.
Pubblicazione: (2025)

Controlled LLM Decoding via Discrete Auto-regressive Biasing
di: Pynadath, Patrick, et al.
Pubblicazione: (2025)

Making Reliable and Flexible Decisions in Long-tailed Classification
di: Li, Bolian, et al.
Pubblicazione: (2025)

SARL: Label-Free Reinforcement Learning by Rewarding Reasoning Topology
di: Wang, Yifan, et al.
Pubblicazione: (2026)

CoT-UQ: Improving Response-wise Uncertainty Quantification in LLMs with Chain-of-Thought
di: Zhang, Boxuan, et al.
Pubblicazione: (2025)

Why Any-Order Autoregressive Models Need Two-Stream Attention: A Structural-Semantic Tradeoff
di: Pynadath, Patrick, et al.
Pubblicazione: (2026)

Boosting Lossless Speculative Decoding via Feature Sampling and Partial Alignment Distillation
di: Gui, Lujun, et al.
Pubblicazione: (2024)

Judge Decoding: Faster Speculative Sampling Requires Going Beyond Model Alignment
di: Bachmann, Gregor, et al.
Pubblicazione: (2025)

Inference-Time Code Selection via Symbolic Equivalence Partitioning
di: Cho, David, et al.
Pubblicazione: (2026)

Generative Frontiers: Why Evaluation Matters for Diffusion Language Models
di: Pynadath, Patrick, et al.
Pubblicazione: (2026)

Learning to Optimize Multi-Objective Alignment Through Dynamic Reward Weighting
di: Lu, Yining, et al.
Pubblicazione: (2025)

MPO: An Efficient Post-Processing Framework for Mixing Diverse Preference Alignment
di: Wang, Tianze, et al.
Pubblicazione: (2025)

Sample-Efficient Alignment for LLMs
di: Liu, Zichen, et al.
Pubblicazione: (2024)

Decoding the Ear: A Framework for Objectifying Expressiveness from Human Preference Through Efficient Alignment
di: Lin, Zhiyu, et al.
Pubblicazione: (2025)

Bayesian Computation in Deep Learning
di: Chen, Wenlong, et al.
Pubblicazione: (2025)

FlashSampling: Fast and Memory-Efficient Exact Sampling
di: Ruiz, Tomas, et al.
Pubblicazione: (2026)

Efficient Alignment of Large Language Models via Data Sampling
di: Khera, Amrit, et al.
Pubblicazione: (2024)

Temper and Tilt Lead to SLOP: Reward Hacking Mitigation with Inference-Time Alignment
di: Wang, Ye, et al.
Pubblicazione: (2026)

Sherlock: Self-Correcting Reasoning in Vision-Language Models
di: Ding, Yi, et al.
Pubblicazione: (2025)

Faster Cascades via Speculative Decoding
di: Narasimhan, Harikrishna, et al.
Pubblicazione: (2024)

Deconvolving Complex Neuronal Networks into Interpretable Task-Specific Connectomes
di: Wang, Yifan, et al.
Pubblicazione: (2024)

VOCABTRIM: Vocabulary Pruning for Efficient Speculative Decoding in LLMs
di: Goel, Raghavv, et al.
Pubblicazione: (2025)

Directional Alignment Mitigates Reward Hacking in Reinforcement Learning for Language Models
di: Deng, Wenlong, et al.
Pubblicazione: (2026)

ARGS: Alignment as Reward-Guided Search
di: Khanov, Maxim, et al.
Pubblicazione: (2024)

Reward-free Alignment for Conflicting Objectives
di: Chen, Peter, et al.
Pubblicazione: (2026)

Transfer Q Star: Principled Decoding for LLM Alignment
di: Chakraborty, Souradip, et al.
Pubblicazione: (2024)

Sample, Don't Search: Rethinking Test-Time Alignment for Language Models
di: Faria, Gonçalo, et al.
Pubblicazione: (2025)

Out-of-Vocabulary Sampling Boosts Speculative Decoding
di: Timor, Nadav, et al.
Pubblicazione: (2025)

Noise Contrastive Alignment of Language Models with Explicit Rewards
di: Chen, Huayu, et al.
Pubblicazione: (2024)

TreeBoN: Enhancing Inference-Time Alignment with Speculative Tree-Search and Best-of-N Sampling
di: Qiu, Jiahao, et al.
Pubblicazione: (2024)

Reward-aware Preference Optimization: A Unified Mathematical Framework for Model Alignment
di: Sun, Shengyang, et al.
Pubblicazione: (2025)