{"total":119,"rows":[{"id":"2112.04359","title":"Ethical and social risks of harm from Language Models","created":"2021-12-08","cited_by_count":966,"area":null},{"id":"2307.15217","title":"Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback","created":"2023-07-27","cited_by_count":759,"area":"alignment"},{"id":"2310.06452","title":"Understanding the Effects of RLHF on LLM Generalisation and Diversity","created":"2023-10-10","cited_by_count":396,"area":"alignment"},{"id":"2302.08582","title":"Pretraining Language Models with Human Preferences","created":"2023-02-16","cited_by_count":247,"area":"alignment"},{"id":"2205.13636","title":"Quark: Controllable Text Generation with Reinforced Unlearning","created":"2022-05-26","cited_by_count":237,"area":"safety"},{"id":"2202.04538","title":"Generating Training Data with Language Models: Towards Zero-Shot Language Understanding","created":"2022-02-09","cited_by_count":199,"area":null},{"id":"2309.16240","title":"Beyond Reverse KL: Generalizing Direct Preference Optimization with Diverse Divergence Constraints","created":"2023-09-28","cited_by_count":174,"area":"alignment"},{"id":"2109.07445","title":"Challenges in Detoxifying Language Models","created":"2021-09-15","cited_by_count":169,"area":null},{"id":"2202.11705","title":"COLD Decoding: Energy-based Constrained Text Generation with Langevin Dynamics","created":"2022-02-23","cited_by_count":161,"area":null},{"id":"2406.16838","title":"From Decoding to Meta-Generation: Inference-time Algorithms for Large Language Models","created":"2024-06-24","cited_by_count":137,"area":null},{"id":"2205.11275","title":"RL with KL penalties is better viewed as Bayesian inference","created":"2022-05-23","cited_by_count":135,"area":null},{"id":"2404.11999","title":"Token-level Direct Preference Optimization","created":"2024-04-18","cited_by_count":135,"area":"alignment"},{"id":"2201.05337","title":"A Survey of Controllable Text Generation using Transformer-based Pre-trained Language Models","created":"2022-01-14","cited_by_count":129,"area":null},{"id":"2201.05273","title":"Pretrained Language Models for Text Generation: A Survey","created":"2022-01-14","cited_by_count":122,"area":null},{"id":"2302.08215","title":"Aligning Language Models with Preferences through f-divergence Minimization","created":"2023-02-16","cited_by_count":120,"area":"alignment"},{"id":"2402.14740","title":"Back to Basics: Revisiting REINFORCE Style Optimization for Learning from Human Feedback in LLMs","created":"2024-02-22","cited_by_count":108,"area":"alignment"},{"id":"2310.04373","title":"Confronting Reward Model Overoptimization with Constrained RLHF","created":"2023-10-06","cited_by_count":96,"area":"alignment"},{"id":"2404.17546","title":"Probabilistic Inference in Language Models via Twisted Sequential Monte Carlo","created":"2024-04-26","cited_by_count":90,"area":"alignment"},{"id":"2206.00761","title":"On Reinforcement Learning and Distribution Matching for Fine-Tuning Language Models with no Catastrophic Forgetting","created":"2022-06-01","cited_by_count":82,"area":null},{"id":"2306.03081","title":"Sequential Monte Carlo Steering of Large Language Models using Probabilistic Programs","created":"2023-06-05","cited_by_count":80,"area":null},{"id":"2107.03451","title":"Anticipating Safety Issues in E2E Conversational AI: Framework and Tooling","created":"2021-07-07","cited_by_count":78,"area":null},{"id":"2203.13299","title":"Mix and Match: Learning-free Controllable Text Generation using Energy Language Models","created":"2022-03-24","cited_by_count":66,"area":null},{"id":"2205.12558","title":"Gradient-Based Constrained Sampling from Language Models","created":"2022-05-25","cited_by_count":64,"area":null},{"id":"2206.08325","title":"Characteristics of Harmful Text: Towards Rigorous Benchmarking of Language Models","created":"2022-06-16","cited_by_count":49,"area":null},{"id":"2306.11816","title":"Learning to Generate Better Than Your LLM","created":"2023-06-20","cited_by_count":49,"area":null},{"id":"2205.14219","title":"Controllable Text Generation with Neurally-Decomposed Oracle","created":"2022-05-27","cited_by_count":41,"area":null},{"id":"2211.03044","title":"Tuning Language Models as Training Data Generators for Augmentation-Enhanced Few-Shot Learning","created":"2022-11-06","cited_by_count":36,"area":null},{"id":"2305.14934","title":"GRACE: Discriminator-Guided Chain-of-Thought Reasoning","created":"2023-05-24","cited_by_count":35,"area":"lr"},{"id":"2210.02889","title":"A Distributional Lens for Multi-Aspect Controllable Text Generation","created":"2022-10-06","cited_by_count":33,"area":null},{"id":"2406.17692","title":"From Distributional to Overton Pluralism: Investigating Large Language Model Alignment","created":"2024-06-25","cited_by_count":33,"area":"alignment"},{"id":"2104.08790","title":"Misinfo Reaction Frames: Reasoning about Readers' Reactions to News Headlines","created":"2021-04-18","cited_by_count":32,"area":null},{"id":"2112.00791","title":"Controlling Conditional Language Models without Catastrophic Forgetting","created":"2021-12-01","cited_by_count":32,"area":null},{"id":"2306.00398","title":"Preference-grounded Token-level Guidance for Language Model Fine-tuning","created":"2023-06-01","cited_by_count":31,"area":"alignment"},{"id":"2406.16235","title":"Preference Tuning For Toxicity Mitigation Generalizes Across Languages","created":"2024-06-23","cited_by_count":31,"area":"alignment"},{"id":"2210.09551","title":"DisCup: Discriminator Cooperative Unlikelihood Prompt-tuning for Controllable Text Generation","created":"2022-10-18","cited_by_count":30,"area":null},{"id":"2211.08714","title":"Reward Gaming in Conditional Text Generation","created":"2022-11-16","cited_by_count":30,"area":null},{"id":"2208.00638","title":"Composable Text Controls in Latent Space with ODEs","created":"2022-08-01","cited_by_count":28,"area":null},{"id":"2209.06970","title":"Generative Visual Prompt: Unifying Distributional Control of Pre-Trained Generative Models","created":"2022-09-14","cited_by_count":27,"area":"diffusion"},{"id":"2212.08307","title":"Controllable Text Generation via Probability Density Estimation in the Latent Space","created":"2022-12-16","cited_by_count":27,"area":null},{"id":"2405.15624","title":"Inverse-RLignment: Large Language Model Alignment from Demonstrations through Inverse Reinforcement Learning","created":"2024-05-24","cited_by_count":27,"area":"alignment"},{"id":"2305.12018","title":"BOLT: Fast Energy-based Controlled Text Generation with Tunable Biases","created":"2023-05-19","cited_by_count":24,"area":null},{"id":"2111.02878","title":"Unsupervised and Distributional Detection of Machine-Generated Text","created":"2021-11-04","cited_by_count":23,"area":null},{"id":"2207.09542","title":"Controllable Data Generation by Deep Learning: A Review","created":"2022-07-19","cited_by_count":20,"area":null},{"id":"2309.09055","title":"Exploring the impact of low-rank adaptation on the performance, efficiency, and regularization of RLHF","created":"2023-09-16","cited_by_count":20,"area":"alignment"},{"id":"2206.05395","title":"Why is constrained neural language generation particularly challenging?","created":"2022-06-11","cited_by_count":19,"area":null},{"id":"2109.05140","title":"Refocusing on Relevance: Personalization in NLG","created":"2021-09-10","cited_by_count":18,"area":null},{"id":"2407.05694","title":"On the Limitations of Compute Thresholds as a Governance Strategy","created":"2024-07-08","cited_by_count":18,"area":null},{"id":"2306.00784","title":"Interpretable Math Word Problem Solution Generation Via Step-by-step Planning","created":"2023-06-01","cited_by_count":17,"area":null},{"id":"2309.05454","title":"Flesch or Fumble? Evaluating Readability Standard Alignment of Instruction-Tuned Language Models","created":"2023-09-11","cited_by_count":17,"area":null},{"id":"2505.11080","title":"BLEUBERI: BLEU is a surprisingly effective reward for instruction following","created":"2025-05-16","cited_by_count":17,"area":"alignment"},{"id":"2403.11558","title":"Reinforcement Learning with Token-level Feedback for Controllable Text Generation","created":"2024-03-18","cited_by_count":16,"area":null},{"id":"2410.12138","title":"Preference Optimization with Multi-Sample Comparisons","created":"2024-10-16","cited_by_count":16,"area":"diffusion"},{"id":"2410.03767","title":"Reasoning Elicitation in Language Models via Counterfactual Feedback","created":"2024-10-02","cited_by_count":14,"area":"lr"},{"id":"2506.06298","title":"Pairwise Calibrated Rewards for Pluralistic Alignment","created":"2025-05-17","cited_by_count":14,"area":"alignment"},{"id":"2403.12017","title":"Supervised Fine-Tuning as Inverse Reinforcement Learning","created":"2024-03-18","cited_by_count":12,"area":"alignment"},{"id":"2501.02790","title":"Segmenting Text and Learning Their Rewards for Improved RLHF in Language Model","created":"2025-01-06","cited_by_count":12,"area":"alignment"},{"id":"2210.10329","title":"Language Detoxification with Attribute-Discriminative Latent Space","created":"2022-10-19","cited_by_count":11,"area":null},{"id":"2308.14328","title":"Reinforcement Learning for Generative AI: A Survey","created":"2023-08-28","cited_by_count":11,"area":null},{"id":"2310.18342","title":"MIRACLE: Towards Personalized Dialogue Generation with Latent-Space Multiple Personal Attribute Control","created":"2023-10-22","cited_by_count":11,"area":null},{"id":"2403.03419","title":"Negating Negatives: Alignment with Human Negative Samples via Distributional Dispreference Optimization","created":"2024-03-06","cited_by_count":11,"area":"alignment"},{"id":"2501.05220","title":"A Novel Approach to Scalable and Automatic Topic-Controlled Question Generation in Education","created":"2025-01-09","cited_by_count":11,"area":null},{"id":"2106.04985","title":"Energy-Based Models for Code Generation under Compilability Constraints","created":"2021-06-09","cited_by_count":10,"area":null},{"id":"2310.16763","title":"SuperHF: Supervised Iterative Learning from Human Feedback","created":"2023-10-25","cited_by_count":10,"area":"alignment"},{"id":"2301.12729","title":"Response-act Guided Reinforced Dialogue Generation for Mental Health Counseling","created":"2023-01-30","cited_by_count":9,"area":null},{"id":"2412.13862","title":"Energy-Based Preference Model Offers Better Offline Alignment than the Bradley-Terry Preference Model","created":"2024-12-18","cited_by_count":9,"area":"alignment"},{"id":"2503.00723","title":"Re-Imagining Multimodal Instruction Tuning: A Representation View","created":"2025-03-02","cited_by_count":9,"area":"mllm"},{"id":"2310.01041","title":"Language Model Decoding as Direct Metrics Optimization","created":"2023-10-02","cited_by_count":8,"area":null},{"id":"2402.02479","title":"BRAIn: Bayesian Reward-conditioned Amortized Inference for natural language generation from feedback","created":"2024-02-04","cited_by_count":7,"area":"alignment"},{"id":"2406.04460","title":"Evaluating the Smooth Control of Attribute Intensity in Text Generation with LLMs","created":"2024-06-06","cited_by_count":7,"area":null},{"id":"2407.21531","title":"Can LLMs \"Reason\" in Music? An Evaluation of LLMs' Capability of Music Understanding and Generation","created":"2024-07-31","cited_by_count":7,"area":"lr"},{"id":"2410.06716","title":"Guaranteed Generation from Large Language Models","created":"2024-10-09","cited_by_count":7,"area":null},{"id":"2410.11677","title":"Understanding Likelihood Over-optimisation in Direct Alignment Algorithms","created":"2024-10-15","cited_by_count":7,"area":"alignment"},{"id":"2211.05750","title":"Nano: Nested Human-in-the-Loop Reward Learning for Few-shot Language Model Control","created":"2022-11-10","cited_by_count":6,"area":"alignment"},{"id":"2312.04510","title":"A Block Metropolis-Hastings Sampler for Controllable Energy-based Text Generation","created":"2023-12-07","cited_by_count":6,"area":null},{"id":"2406.09688","title":"FreeCtrl: Constructing Control Centers with Feedforward Layers for Learning-Free Controllable Text Generation","created":"2024-06-14","cited_by_count":6,"area":null},{"id":"2503.17482","title":"What's Producible May Not Be Reachable: Measuring the Steerability of Generative Models","created":"2025-03-21","cited_by_count":6,"area":null},{"id":"2510.10020","title":"Calibrating Generative Models to Distributional Constraints","created":"2025-10-11","cited_by_count":6,"area":null},{"id":"2106.03831","title":"Counterfactual Maximum Likelihood Estimation for Training Deep Networks","created":"2021-06-07","cited_by_count":5,"area":null},{"id":"2206.10103","title":"Automatic Controllable Product Copywriting for E-Commerce","created":"2022-06-21","cited_by_count":5,"area":null},{"id":"2412.07334","title":"Frame Representation Hypothesis: Multi-Token LLM Interpretability and Concept-Guided Text Generation","created":"2024-12-10","cited_by_count":5,"area":null},{"id":"2107.13077","title":"Neural Rule-Execution Tracking Machine For Transformer-Based Text Generation","created":"2021-07-27","cited_by_count":4,"area":null},{"id":"2303.05431","title":"disco: a toolkit for Distributional Control of Generative Models","created":"2023-03-08","cited_by_count":3,"area":null},{"id":"2309.12551","title":"Is it Possible to Modify Text to a Target Readability Level? An Initial Investigation Using Zero-Shot Large Language Models","created":"2023-09-22","cited_by_count":3,"area":null},{"id":"2404.07117","title":"Continuous Language Model Interpolation for Dynamic and Controllable Text Generation","created":"2024-04-10","cited_by_count":3,"area":null},{"id":"2406.11039","title":"Dynamic Normativity: Necessary and Sufficient Conditions for Value Alignment","created":"2024-06-16","cited_by_count":3,"area":"alignment"},{"id":"2407.11449","title":"Controllable Contextualized Image Captioning: Directing the Visual Narrative through User-Defined Highlights","created":"2024-07-16","cited_by_count":3,"area":null},{"id":"2512.05962","title":"Whatever Remains Must Be True: Filtering Drives Reasoning in LLMs, Shaping Diversity","created":"2025-12-05","cited_by_count":3,"area":"lr"},{"id":"2112.05702","title":"Sampling from Discrete Energy-Based Models with Quality/Efficiency Trade-offs","created":"2021-12-10","cited_by_count":2,"area":null},{"id":"2211.14540","title":"Lexical Complexity Controlled Sentence Generation","created":"2022-11-26","cited_by_count":2,"area":null},{"id":"2311.07430","title":"Controlled Text Generation for Black-box Language Models via Score-based Progressive Editor","created":"2023-11-13","cited_by_count":2,"area":null},{"id":"2410.00508","title":"FlipGuard: Defending Preference Alignment against Update Regression with Constrained Optimization","created":"2024-10-01","cited_by_count":2,"area":"alignment"},{"id":"2502.10411","title":"TrueReason: An Exemplar Personalised Learning System Integrating Reasoning with Foundational Models","created":"2025-01-23","cited_by_count":2,"area":"lr"},{"id":"2512.16565","title":"Non-Asymptotic Global Convergence of PPO-Clip","created":"2025-12-18","cited_by_count":2,"area":null},{"id":"2602.01685","title":"Semantic-aware Wasserstein Policy Regularization for Large Language Model Alignment","created":"2026-02-02","cited_by_count":2,"area":"alignment"},{"id":"2606.21550","title":"AI Alignment From Social Choice Perspectives","created":"2026-06-19","cited_by_count":2,"area":"alignment"},{"id":"2404.12150","title":"Aligning language models with human preferences","created":"2024-04-18","cited_by_count":1,"area":"alignment"},{"id":"2406.13660","title":"Towards Minimal Targeted Updates of Language Models with Targeted Negative Training","created":"2024-06-19","cited_by_count":1,"area":null},{"id":"2412.18673","title":"MapExplorer: New Content Generation from Low-Dimensional Visualizations","created":"2024-12-24","cited_by_count":1,"area":null},{"id":"2502.05773","title":"PIPA: Preference Alignment as Prior-Informed Statistical Estimation","created":"2025-02-09","cited_by_count":1,"area":"alignment"},{"id":"2502.18644","title":"Steered Generation via Gradient Descent on Sparse Features","created":"2025-02-25","cited_by_count":1,"area":null}],"next_cursor":"100"}