Constrained Decoding

Follow this topic Ask about this topic

Momentum

5 papers in the last four weeks, up 25% on the four weeks before. 0.1% of all new papers.

Jul 6Week of Sep 21

Latest papers 56

All topics

GrammarRL: Effective Grammar-Constrained Decoding via Reinforcement Learning
Sep 30, 2026Gabriele Tuccio, Antonino Furnari, Aldo Gangemi +1Language Modeling Constrained Decoding
University of Catania · ISTC - National Research Council, Italy · University of Bologna
GroundAnything: Reconciling Parallel Decoding with Precise Visual Grounding at Flash Speed
Sep 30, 2026Qize Yu, Lianrui Fan, Bowen Ping +19Weak Visual Grounding Grounding
XPeng Inc. · Peking University · The University of Hong Kong +4
Self-Repulsive Sampling for Diffusion Language Models
Sep 30, 2026Michael Helcig, Martin JaggiDiffusion Language Models Diffusion Sampling
ETH Zürich · EPFL
Fine-Tuning Diffusion Language Models with Context Selection and Target Weighting
Sep 29, 2026Loay Mualem, Lluís Pastor-Pérez, Vinh Tong +4Diffusion Language Models Language Modeling
IMPRS-IS · Institute for AI, University of Stuttgart · Bitdefender
Repetition, Not Length: Isolating the Counting Failure in Neural Text-to-Speech
Sep 29, 2026Kirill Borodin, Vasilii Kudryavtsev, Maxim Maslov +1Repetition Constrained Decoding
BitmanagerAI, Dubai, UAE · lab260, Yerevan, Armenia · MTUCI, Moscow, Russia
Reliable Parallel Decoding in Masked Diffusion Language Models
Sep 29, 2026Zhenghao He, Bohan Liu, Guangzhi Xiong +1Masked Diffusion Language Models Constrained Decoding
Department of Computer Science, University of Virginia
Draft in Parallel, Condition Through Depth: Adjacent Causal Injection for Speculative Decoding
Sep 28, 2026Haohui Zhang, Keyu Chen, Haocheng Sun +4Speculative Decoding Drafter
Shanghai Jiao Tong University · Tencent YouTu Lab · Xiamen University
Twist, Don't Tilt: Trajectory-Exact Constrained Decoding for Masked Diffusion Models
Sep 28, 2026Aditya Thimmaiah, Lara Marinov, Jayanth Srinivasa +3Masked Diffusion Language Models Constrained Decoding
The University of Texas at Austin · Cisco Research
From Position Risks to Block Survival: Faster Generation for Diffusion Language Models
Sep 27, 2026Siwei Chen, Yuxiang Wan, Yifan Yu +1Diffusion Language Models Diffusion Models
University of Illinois Urbana-Champaign
FLEET: From Logits Entropy to Enhanced Trajectories in Text Generation
Sep 23, 2026Oleksii Streltsov, Oleksandra VitkoText Generation Constrained Decoding
Department of Artificial Intelligence, Kharkiv National University of Radio Electronics, Kharkiv, Ukraine
Greedy Decoding Is Not Precision-Invariant: Cross-Precision Output Divergence in LLM Inference
Sep 22, 2026Gaoyuan Du, Anam Nawaz Khan, Rex Zhou +4Constrained Decoding LLM Inference Optimization
University of Tennessee, Knoxville · University of Chicago · Amazon +1
Constrained Decoding Eliminates Structural Failures in Small LLMs but Reveals a Scale-Dependent Semantic Gap
Sep 20, 2026Akash ChavanLarge Language Models Fail Constrained Decoding
Temporal Self-Distillation: Faster Inference in Discrete Diffusion Language Models
Sep 14, 2026Shijian Xu, Andrea Miele, Metod Jazbec +3Diffusion Language Models Fast Inference
University of Basel · University of Amsterdam · Johns Hopkins University
Trie Automata for Constrained Decoding over Large Finite Sets
Aug 12, 2026Xingzi Xu, Karim BouyarmaneConstrained Decoding Context-Free Grammars
Amazon
Actions Speak Louder than Words: Measuring Cross-Lingual Policy Retention in Tool-Using Agents
Aug 11, 2026Sourabrata Mukherjee, Kalika Bali, Sunayana SitaramMultilingual Agents Constrained Decoding
Microsoft Research India
The Parser Already Knows: Lightweight Bias Correction in Constrained Decoding
Aug 10, 2026Işıl Özgü, Yaoxuan Wu, Guy Van den Broeck +1Constrained Decoding Correction
University of California, Los Angeles
Screenshots or Tools? Eliciting Tool Use and Managing Multimodal Context in Hybrid GUI-MCP Computer-Use Agents
Aug 4, 2026Siqi Fan, Minghao Li, Xiaoqian Ma +6Computer-Use Agents Multimodal Agents
University of Electronic Science and Technology of China · AI Platform, Xiaohongshu Inc. · Gaoling School of Artificial Intelligence, Renmin University of China +1
Celty: SpMspV GPU Kernel and SIMT Co-Design for Efficient Dual-Sparse LLM Inference
Aug 2, 2026Ruokai Yin, Priyadarshini PandaLLM Inference Optimization Dynamic Sparse Attention
Yale University New Haven, USA · University of Southern California Los Angeles, USA
Escaping Confidence Trap: Evolutionary Decoding for Mathematical Reasoning in Diffusion LLMs
Aug 1, 2026Zhenhong Sun, Hanqing Zhao, Yatao Bian +7Mathematical Reasoning Diffusion Language Models
Australian National University Canberra, Australia · Nanyang Technological University Singapore · National University of Singapore Singapore +2
Stabilized Best-of- $K$ Training for Neural Combinatorial Optimization
Jul 31, 2026Melveena Jolly, Midhun XavierNeural Combinatorial Optimization Reward Signal
1Independent Researcher
Explicit Note-Event Tokenization and Pitch-Validity Constrained Decoding for MIDI-to-Tablature Transcription
Jul 29, 2026Ting-Kai Hsu, Wei-Chin Wang, Kai-Xi Hong +1Multi-Instrument Music Transcription Transcription
Graduate Institute of Communication Engineering National Taiwan University Taipei, Taiwan · Department of Electrical Engineering National Taiwan University Taipei, Taiwan · Graduate Institute of Networking and Multimedia National Taiwan University Taipei, Taiwan
From Evaluation to Optimisation: Hierarchy-Aware Training Signals for CWE Prediction in Python
Jul 23, 2026Muntasir Adnan, Manile Srun, Carlos C. N. KuhnQuadratic Slack Penalty Destination-Only Scoring
Open Source Institute, Faculty of Science and Technology, University of Canberra, Canberra, Australia.
HPD-Parsing: Hierarchical Parallel Document Parsing
Jul 21, 2026Shu Wei, Jingjing Wu, Lingshu Zhang +10Document Parsing Multi-Token Prediction
Policy-Conditioned Constrained Decoding for Column-Level Access Control in Text-to-SQL
Jul 14, 2026Ryoto Miyamoto, Xin Fan, Hayato YamanaText-To-Sql Access Control
Waseda University
Hallucination Detection in Large Language Models Using Diversion Decoding
Jul 11, 2026Basel Abdeen, S M Tahmid Siddiqui, Meah Tahmeed Ahmed +4Large Language Model Hallucination Hallucination Detection
The University of Texas at Dallas, Richardson, TX 75080, USA · National Institute of Standards and Technology, Gaithersburg, USA · Carnegie Mellon University, PA, USA
FlashTrie: A GPU-Accelerated Constrained Beam Search for Generative Retrieval
Jul 10, 2026Dakshitha Anandakumar, Anurag Mukkara, Wenxiang Hu +5Beam Search Constrained Decoding
Microsoft, Redmond, WA, USA · Nvidia, Santa Clara, CA, USA
Constrained Decoding for Diffusion Language Models via Efficient Inference over Finite Automata
Jul 8, 2026Meihua Dang, Stefano ErmonConstrained Decoding Diffusion Language Models
Stanford University
Mitigating Errors in LLM-Generated Web API Invocations via Retrieval-Augmented Generation and Constrained Decoding
Jul 7, 2026Daniel Maninger, Leon Chemnitz, Jannis Brugger +3Code Generation Constrained Decoding
Technische Universität Darmstadt, Darmstadt, Germany · Hessian Center for Artificial Intelligence (hessian.AI), Darmstadt, Germany · Pariton AI, Berlin, Germany +1
Noisy-Channel Minimum Bayes Risk Decoding
Jul 6, 2026Yusuke Sakai, Hidetaka Kamigaito, Taro WatanabeConstrained Decoding Hidden Markov Models
Nara Institute of Science and Technology, Nara, Japan.
The Anatomy of the CTC Oracle Gap: Acoustic Exhaustion and Linguistic Recovery
Jun 22, 2026Ivan NovosadConstrained Decoding Linguistics
Faculty of Computer Science, HSE University, Moscow, Russia.
Who Should Lead Decoding Now? Tracking Reliable Trajectories for Ensembling Masked Diffusion Language Models
Jun 15, 2026Heecheol Yun, Joonhyung Park, Joowon Kim +1Masked Diffusion Language Models Constrained Decoding
1KAIST · 2AITRICS
Mean-Field Parallel Decoding for Discrete Diffusion Language Models
Jun 14, 2026Tamim Zoabi, Ameen Ali, Liran Ringel +1Diffusion Language Models Constrained Decoding
School of Electrical & Computer Engineering Tel Aviv University, Israel · School of Computer Science and AI Tel Aviv University, Israel · Department of Computer Science Technion, Israel Institute of Technology, Israel
Express Language Modeling
Jun 9, 2026Albert Gong, Annabelle Michael Carrell, Raaz Dwivedi +1Language Modeling Causal Attention
Cornell Tech · University of Cambridge · Microsoft Research
Mitigating Bias in Locally Constrained Decoding via Tractable Proposals
Jun 1, 2026Meihua Dang, Linxin Song, Honghua Zhang +3Constrained Decoding Large Language Model Generation
Department of Computer Science, Stanford University · Department of Computer Science, University of Southern California · Department of Computer Science, University of California, Los Angeles
DSL-LLaDA: Scaling Continuous Denoising to 8B Masked Diffusion LMs
May 31, 2026Longxuan Yu, Yunshu Wu, Yu Fu +5Masked Diffusion Language Models Diffusion Models
University of California, Riverside · Georgia Institute of Technology · Microsoft
EPIC: Efficient and Parallel Inference under CFG Constraints for Diffusion Language Models
May 30, 2026Hyundong Jin, Yo-Sub HanConstrained Decoding Diffusion Language Models
Yonsei University, Seoul, Republic of Korea
Efficient Diffusion LLMs via Temporal-Spatial Parallel Decoding and Confidence Extrapolation
May 29, 2026Zekai Li, Ji Liu, Yiqing Huang +3Diffusion Language Models Constrained Decoding
Advanced Micro Devices, Inc. (AMD)
BlockBatch: Multi-Scale Consensus Decoding for Efficient Diffusion Language Model Inference
May 28, 2026Xiaoyou Wu, Cheng-Jhih Shih, Binfei Ji +2Diffusion Language Models LLM Inference Optimization
Georgia Institute of Technology
The Confidence Shortcut: A Reasoning Failure Mode of Masked Diffusion Models
May 27, 2026Dueun Kim, Albert NoMasked Diffusion Language Models Constrained Decoding
Department of Artificial Intelligence, Yonsei University
Diffusion Large Language Models for Visual Speech Recognition
May 27, 2026Jeong Hun Yeo, Chae Won Kim, Hyeongseop Rha +1Diffusion Language Models Constrained Decoding
Integrated Vision Language Lab, KAIST, South Korea
Dynamic-dLLM: Dynamic Cache-Budget and Adaptive Parallel Decoding for Training-Free Acceleration of Diffusion LLM
May 27, 2026Tianyi Wu, Xiaoxi Sun, Yanhua Jiao +5Diffusion Language Models LLM Inference Optimization
*Equal contribution. The work is done during Tianyi’s internship at CBG Celia DeviceAI Team, Huawei. · 1Harbin Institute of Technology, Shenzhen · 2Huawei +1
LocateAnything: Fast and High-Quality Vision-Language Grounding with Parallel Box Decoding
May 26, 2026Shihao Wang, Shilong Liu, Yuanguo Kuang +10Vision-Language Model Grounding Weak Visual Grounding
The Hong Kong Polytechnic University · NVIDIA · Princeton University +2
Maximum Likelihood Decoding of Quantum Error Correction Codes
May 17, 2026Hanyan Cao, Ge Yan, Yuxuan Du +1Quantum Error Correction Post-Quantum Cryptography
1Science, Mathematics and Technology Cluster, Singapore University of Technology and Design, 8 Somapah Road, 487372, Singapore · College of Computing and Data Science, Nanyang Technological University, Singapore, Singapore · School of Physical and Mathematical Sciences, Nanyang Technological University,2026 Singapore, Singapore
Task-Aware Calibration: Provably Optimal Decoding in LLMs
May 11, 2026Tim Tomov, Dominik Fuchsgruber, Rajeev Verma +1Constrained Decoding
School of Computation, Information & Technology, Technical University of Munich · Munich Data Science Institute · Munich Center for Machine Learning +1
NCO: A Versatile Plug-in for Handling Negative Constraints in Decoding
May 11, 2026Hyundong Jin, Yo-Sub HanConstrained Decoding Large Language Model Generation
Yonsei University Seoul, Republic of Korea
Entropy-informed Decoding: Adaptive Information-Driven Branching
May 10, 2026Benjamin Patrick Evans, Sumitra Ganesh, Leo ArdonConstrained Decoding Token-Level Entropy
JP Morgan AI Research, London, UK · JP Morgan AI Research, New York, USA
Reliable Answers for Recurring Questions: Boosting Text-to-SQL Accuracy with Template Constrained Decoding
Apr 30, 2026Smit Jivani, Sarvam Maheshwari, Sunita SarawagiText-To-Sql Constrained Decoding
Department of Computer Science and Engineering, Indian Institute of Technology Bombay, India
Revisiting Greedy Decoding for Visual Question Answering: A Calibration Perspective
Apr 25, 2026Boqi Chen, Xudong Liu, Yunke Ao +1Constrained Decoding Answer
1ETH Zurich · University of Toronto · 3MBZUAI
Lost in Decoding? Reproducing and Stress-Testing the Look-Ahead Prior in Generative Retrieval
Apr 25, 2026Kidist Amde Mekonnen, Yongkang Li, Yubao Tang +2Beam Search Constrained Decoding
University of Amsterdam Amsterdam, The Netherlands
LatticeBridge: Rare-Event Sequential Inference for Faithful Structured Sequence Synthesis
Apr 22, 2026Faruk Alpay, Bugra KilictasSequence Modeling Constrained Decoding
Department of Computer Engineering, Bahcesehir University
The Query Channel: Information-Theoretic Limits of Masking-Based Explanations
Apr 17, 2026Erciyes Karakaya, Ozgur ErcetinInformation-Theoretic Limits Explainable AI Methods
*Department of Electrical and Computer Engineering, University of Maryland, College Park, USA · †Faculty of Engineering and Natural Sciences, Sabanci University, Turkiye
The Illusion of Equivalence: Systematic FP16 Divergence in KV-Cached Autoregressive Inference
Apr 16, 2026Ranjith Chodavarapu, Lei XuMatched Fp16 Intermediate Autoregressive Transformers
Department of Computer Science Kent State University, USA
Schema-Key Wording as an Instruction Channel in Structured Generation under Constrained Decoding
Apr 16, 2026Yifan LeConstrained Decoding Data Generation
Independent Researcher
Few-Step Diffusion Language Models via Trajectory Self-Distillation
Feb 12, 2026Tunyu Zhang, Xinxi Zhang, Ligong Han +10Diffusion Language Models Few-Step Generation
Rutgers University · Red Hat AI Innovation · MIT-IBM Watson AI Lab
The Hidden Cost of Structured Generation in LLMs: Draft-Conditioned Constrained Decoding
Feb 8, 2026Avinash Reddy, Thayne T. Walker, James S. Ide +1Constrained Decoding Large Language Model Generation
Department of Computer Science, University of Central Florida, FL, US · Lockheed Martin AI Center, CT, US
Grammar-Aligned Decoding
May 31, 2024Kanghee Park, Jiayu Wang, Taylor Berg-Kirkpatrick +2Constrained Decoding Large Language Model Generation
University of Wisconsin-Madison · University of California San Diego

Sep 30, 2026, cs.AIJ/K move · Enter open · S save

GrammarRL: Effective Grammar-Constrained Decoding via Reinforcement Learning

Gabriele Tuccio, Antonino Furnari, Aldo Gangemi, Misael Mongiovì

University of Catania · ISTC - National Research Council, Italy · University of Bologna

Language Modeling Constrained Decoding Large Language Model Reinforcement Learning Grammaticality Beam Search Decoding

Open paper AskPDF arXiv

Grammar-constrained generation guarantees syntactic validity, but can substantially degrade semantic quality when the model's preferred outputs are poorly aligned with the imposed grammar. This trade-off is particularly severe when the prompt is underspecified or the model has limited instruction-following ability. Beam search can partially mitigate these failures by exploring multiple valid sequences, but its computational cost grows with beam width, while sequence-level probability is only an imperfect proxy for semantic quality. We introduce GrammarRL, a label-free reinforcement learning method that adapts language models to grammar constraints without requiring annotated data. GrammarRL optimizes the model using two complementary self-supervised rewards derived from its own likelihoods: a direct reward, measuring how likely the constrained output is given the input, and a reverse reward, measuring how well the input can be reconstructed from the generated output. We optimize these rewards with a Reinforce Leave-One-Out (RLOO) objective over groups of grammar-constrained rollouts, augmented with the top-1 beam-search hypothesis and regularized towards a frozen base model. We evaluate GrammarRL on sign language gloss translation, hierarchical text classification, and named entity recognition using Llama models ranging from 1B to 8B parameters. GrammarRL consistently outperforms constrained greedy decoding, with an average improvement of 9.8 points and gains of up to 22.8 BLEU. It matches or outperforms beam search on two of the three tasks while preserving greedy-decoding inference cost. Ablations further show that the two rewards are complementary: either reward alone can underperform the untrained baseline, whereas their combination consistently improves upon it.

Momentum

Related topics

Latest papers 56