Research
On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

168,695 papers · 148 categories

Trend · papers per month

4079119158 · May 202619922001200920172026
48 results for reward alignment

Extends reinforcement learning alignment to scalar rewards, improving math reasoning.

problem Designing reinforcement learning algorithms for general LLM alignment.
method Introduces f-GRPO and f-HAL, estimating f-divergences between reward-aligned and unaligned distributions.
result Improves math reasoning RLVR tasks and mitigates reward hacking.

Align-RUDDER improves reinforcement learning with few demonstrations by redistributing rewards.

problem Learning complex tasks with sparse and delayed rewards using few demonstrations.
method Align-RUDDER uses a profile model for reward redistribution based on multiple sequence alignment of demonstrations.
result Align-RUDDER outperforms competitors on complex tasks with few demonstrations.

The paper analyzes the reward improvement of aligned policies in large language models.

problem Optimizing policies in large language models while staying close to a reference policy.
method Information-theoretic analysis and reduction to exponential order statistics.
result Information-theoretic upper bounds on reward improvement are derived.

PILAF optimizes reward models from human feedback for better policy alignment.

problem Creating accurate reward models from human feedback for policy optimization.
method Policy-Interpolated Learning for Aligned Feedback (PILAF) that explicitly aligns preference learning with maximizing underlying oracle reward.
result PILAF is optimal from both optimization and statistical perspectives, demonstrating strong performance in RLHF settings.

EBRM improves robustness and generalization of language model rewards.

problem Challenges in capturing complex human preferences and generalizing to unseen data in reward models.
method Energy-Based Reward Model (EBRM) that models reward distribution explicitly, using conflict-aware data filtering, label-noise-aware contrastive training, and hybrid initialization.
result Significant improvements in robustness and generalization, up to 5.97% improvement in safety-critical alignment tasks.

Paper addresses reward hacking in preference optimization, proposing POWER-DL to improve AI alignment.

problem Reward hacking problem in preference optimization, leading to undesired behaviors.
method POWER-DL combines robust reward maximization and dynamic label updates to mitigate reward hacking.
result POWER-DL consistently outperforms state-of-the-art methods on alignment benchmarks.

Novel alignment framework for text-to-image generation using diffusion models and flow matching.

problem Improving text-to-image generation with minimal computational resources.
method Proposes a novel alignment framework that decomposes the score function into pre-trained score plus a conditional expectation of the reward.
result Achieves comparable performance to finetuning-based models with reduced computational cost.

One obstacle to applying reinforcement learning algorithms to real-world problems is the lack of suitable reward functions. Designing such reward functions is difficult in part because the user only has an implicit understanding of the task objective. This gives rise to the agent alignment problem: how do we create age…

2018-11-19abs ↗pdf ↗

SARA uses similarity to learn rewards robustly and adaptively.

problem Robustness to labeler errors and adaptability to diverse feedback formats.
method Contrastive framework that learns latent representations and computes rewards as similarities.
result Strong performance on offline RL benchmarks and diverse applications.

DPA aligns LLMs with multi-objective rewards for diverse user preferences.

problem Fine-grained control over LLMs for diverse user needs.
method Integrates multi-objective reward modeling and directional preference control.
result DPA offers better performance trade-offs and intuitive user control over LLM generation.

Unified framework for aligning and composing diffusion models to satisfy multiple constraints.

problem Improving quality and compliance of generated samples from diffusion models.
method Constrained optimization framework that unifies alignment and composition of diffusion models.
result Proposed framework effectively satisfies multiple constraints in image generation.

RAFT fine-tunes models using high-quality samples to align them with human preferences.

problem Aligning generative models with human ethics and preferences.
method RAFT selects high-quality samples, discards undesired behavior, and fine-tunes the model on filtered samples.
result RAFT improves model performance in reward learning and automated metrics.

Method tackles uncertainty in reward models for LLMs from heterogeneous human feedback.

problem Uncertainty in reward models for LLMs from heterogeneous human feedback.
method Heterogeneous preference framework and alternating gradient descent algorithm.
result Established theoretical guarantees for estimator convergence and asymptotic distribution.

MFMs enable efficient reward alignment for generative models.

problem Computational bottleneck in controlling generative models.
method Meta Flow Maps (MFMs) extend consistency models and flow maps to stochastic regime for efficient value function estimation.
result MFMs enable inference-time steering and unbiased, off-policy fine-tuning to general rewards efficiently.

New method prevents RLHF alignment collapse by accounting for policy's influence on reward model updates.

problem Iterative RLHF leads to alignment collapse where policies exploit RM's blind spots.
method Foresighted policy optimization (FPO) restores missing steering term via regularization.
result FPO prevents alignment collapse on LLM alignment pipelines using Llama-3.2-1B.

New algorithm improves inference-time alignment without reward hacking.

problem Improving quality of responses from language models with limited compute.
method Inference-time alignment, focusing on extttInferenceTimePessimism exttt{InferenceTimePessimism} algorithm.
result Optimal performance and scaling-monotonicity of extttInferenceTimePessimism exttt{InferenceTimePessimism}.

Paper explores limits and possibilities of aligning LLMs with human preferences.

problem Aligning LLMs with diverse human preferences to ensure fairness and informed outcomes.
method Analysis of probabilistic representation of human preferences and preservation of diverse preferences.
result LLMs can't fully align with human preferences using reward-based approaches due to Condorcet cycles, but mixed strategies are statistically possible.

Mitigates biases in reward models using variational inference.

problem Spurious correlations in reward models that align large language models with human preferences.
method Formulates data-generating process, identifies non-spurious latent variables, and uses variational inference to recover them.
result Effective mitigation of spurious correlation issues, yielding more robust reward models.

New method optimizes policies without assuming known link functions between preferences and rewards.

problem Policy alignment with unknown and unrestricted link functions.
method Formulates an ff-divergence-constrained reward maximization problem, learning policies directly.
result Induces a semiparametric single-index binary choice model for policy alignment.

RePULSe improves language model alignment by reducing undesired outputs without sacrificing overall performance.

problem Aligning language models with human preferences while minimizing undesired outputs.
method Integrates probabilistic inference into RL training to reduce undesired outputs.
result RePULSe achieves a better balance between expected reward and undesired output probability.

DRO-REBEL improves LLM alignment by robustly updating models online.

problem Overfitting and drifting of LLMs during RLHF.
method DRO-REBEL uses type-pp Wasserstein, KL, and χ2χ^2 ambiguity sets for robust online updates.
result DRO-REBEL achieves faster convergence and better performance than prior methods.

DTS improves inference-time alignment of diffusion models with less compute.

problem Inference-time alignment of diffusion models suffers from inaccurate value estimation and inefficient reuse of past computations.
method Diffusion Tree Sampling (DTS) uses a tree-based approach to propagate terminal rewards and iteratively refine value estimates.
result DTS produces asymptotically exact samples and matches the FID of best-performing baselines with up to 10x less compute.

Trade-R1 bridges verifiable rewards to stochastic financial markets via process-level reasoning verification.

problem Extending RL to financial markets where rewards are verifiable but noisy.
method A verification method that transforms reasoning over financial documents into a structured RAG task, using a triangular consistency metric.
result DSR achieves superior cross-market generalization while maintaining reasoning consistency.

Reward collapse occurs when ranking-based reward models yield uniform rewards for different prompts.

problem Reward collapse in aligning large language models with human preferences.
method Introduced a prompt-aware optimization scheme to derive closed-form expressions for reward distributions.
result Our prompt-aware utility functions significantly alleviate reward collapse during training.

Improves inference-time alignment for diffusion models without updating weights.

problem Aligning diffusion models without updating weights for high-reward outputs.
method Trust-Region Iterative Twisted Sequential Monte Carlo (TRI-TSMC) for variance reduction and efficiency.
result Improves primary alignment objectives on text generation tasks.

QAlign improves language model alignment with less compute, outperforming existing methods.

problem Improving language model performance with limited test-time computation.
method QAlign: sampling from optimal aligned distribution using Markov chain Monte Carlo.
result Consistent improvements over existing methods on various benchmarks.

A new method learns continuous guidance weights to improve diffusion model quality and distributional alignment.

problem Improving perceptual quality and distributional alignment of samples from conditional diffusion models.
method Learned continuous guidance weights ωc,(s,t)ω_{c,(s,t)} are used to minimize distributional mismatch and reward guided sampling.
result Improvements in Fréchet inception distance (FID) for image generation and better image-prompt alignment in text-to-image applications.

Mitigates overoptimization in RLHF by reformulating SFT loss as a preference optimization loss.

problem Overoptimization in RLHF where reward model misguides generative model.
method Proposes a theoretical algorithm that minimizes maximum likelihood estimation and reward penalty, reformulates as simple objective combining preference optimization and supervised learning losses.
result Improved performance of RPO compared to DPO baselines in aligning LLMs.

New RLHF approach mitigates bias in aligning LLMs with human preferences.

problem Algorithmic bias in RLHF leading to preference collapse.
method Preference Matching (PM) RLHF, using PM regularizer and conditional variant.
result 29% to 41% improvement in alignment with human preferences.

Improved inference-time alignment using Best-of-N and smoothing.

problem Reward overoptimization in Best-of-N (BoN) due to poor proxy reward models.
method Introduced Soft Best-of-N (SBoN) and analyzed its performance through KL divergence and regret analysis.
result Smoothing helps SBoN mitigate reward overoptimization, especially when proxy reward quality is low.

The abstract explores connections between reinforcement learning, scaling, and diffusion.

problem Aligning reinforcement learning with human feedback and scaling techniques.
method Clarifying connections between reinforcement learning, scaling, and diffusion.
result Introducing a resampling approach for alignment and reward-directed diffusion models.

RLHF uses human feedback to train AI models, posing statistical challenges.

problem Aligning AI models with human preferences using noisy, subjective feedback.
method Supervised fine-tuning, reward modeling, policy optimization, statistical ideas.
result Statistical methods for reward function learning and policy optimization.

This paper introduces ff-DPO, a generalized approach to Direct Preference Optimization using diverse divergence constraints.

problem Aligning large language models with human preferences while mitigating safety risks.
method Incorporates diverse divergence constraints to simplify the relationship between reward and optimal policy, eliminating the need for estimating the normalizing constant.
result Optimizes LLMs to align with human preferences more efficiently and under a broader set of divergence constraints.

Dual active learning improves RLHF by selecting optimal conversations and teachers.

problem Efficiently aligning LLMs with human preferences using RLHF from feedback.
method Offline RL for conversation and teacher selection, dual active reward learning, pessimistic RL.
result The proposed algorithm achieves minimal generalized variance and outperforms state-of-the-arts.

Paper resolves bias in ALFT training using generalized alignment games.

problem Systematic bias in estimating logarithmic rewards from small batches.
method Generalized Distributional Alignment Games, U-statistics, minimax polynomial estimators, Variance-Optimal Augmented Polynomial Optimization Program (AQP) Estimator.
result Proves optimal bias and accelerated convergence in ALFT training.

Improved RL training for DMs reduces mode collapse and preserves diversity.

problem Mode collapse and training instability in RL fine-tuned diffusion models.
method Dynamic hierarchical RL training with sliding-window parameter regularisation.
result Models trained with HRF achieve better preservation of diversity in downstream tasks.

Proposes a new RL method to fine-tune flow-based models with arbitrary rewards.

problem Challenges in fine-tuning continuous flow-based generative models with arbitrary reward functions.
method Online Reward-Weighted Conditional Flow Matching with Wasserstein-2 Regularization (ORW-CFM-W2)
result Achieves optimal policy convergence with controllable trade-offs between reward maximization and diversity preservation.