TodayNewsResearchFieldsStacksMarketAnalytics
⌘K

Research

New papers ranked by attention, with their code, models and discussion linked. Pick a field on the left.

WeekMonthQuarterAll
NewestMost relevant

Papers

Open problems

Rising research topics

Ask the research

Sample
ByteTrail · AI and developer-tools intelligence
NewsResearchStacksAPI
Fields · this week
All fields2047Multimodal338Frontier LLMs255Recursive self improvement241Agents & indirect prompt injection189Inference & serving184Neural operators159Clinical reasoning145Robotics127Forecasting time82Linear bandits56AI safety51Voice AI44Character recognition OCR30Molecular property prediction28Federated learning22All 23 fields

Open problems across fields

  • Understanding and adapting to evolving implicit user intent1 paper raises it · 1 addresses it

    Current models fail to infer and continuously update implicit user intents, limiting real-world interactive applications.

    • StreamVibe: Proactive Video Agents for Evolving User Intentraises · addresses
  • Trustworthy recursive self-improvement2 papers raise it · 2 address it

    Self-modifying systems risk flawed evaluations, capability regression, constraint violations, and security or privacy breaches without proper safeguards.

    • Trustworthy Recursive Self-Improvement in the Era of Large Models: Foundation, Risks, Evaluation and Open Challengesraises · addresses
    • The Trinity:Capability-Driven Recursive Self-Improvement for Multimodal Agentsraises · addresses
  • Token-level credit assignment in diffusion language models1 paper raises it · 1 addresses it

    Iterative denoising in diffusion models obscures token contributions, making reinforcement learning signals noisy and ineffective.

    • DTCA: Denoising Token Credit Assignment for Reinforcement Learning in Diffusion Language Modelsraises · addresses
  • Physical dynamics understanding in vision-language-action models1 paper raises it · 1 addresses it

    VLA models lack robust physical reasoning, limiting their effectiveness in real-world robotics tasks.

    • PDRVLA: A Vision-Language-Action Model with Physical Dynamics Representationsraises · addresses
  • Rigorous significance testing for temporal clinical hypotheses1 paper raises it · 1 addresses it

    Language models often assert temporal relationships without statistical validation, risking incorrect conclusions in high-stakes domains like healthcare.

    • Testimony from Timelines: Evidence-Grounded Significance Testing of Temporal Clinical Hypothesesraises · addresses

Named by an AI from these papers' abstracts.

Rising research topics

  • Spiking neural networks7+40.0%
  • Algorithmic fairness9+28.6%
  • Diffusion priors16+14.3%
  • Quantum machine learning16−11.1%
  • Linear bandits56−20.0%
  • Anomaly detectors17−26.1%
  • Character recognition OCR30−26.8%
  • Channel state information8−27.3%

Papers

  • arXiv cs.CLOct 8

    Local Prototype Reconstruction for Text-Compatible Speech-to-LLM Bridge Pretraining

    Xinnian Zhao, Chia-Hua Wu, Pu Wang, Hugo Van Hamme

    Speech-to-LLM systems often connect a frozen speech encoder to a frozen large language model (LLM) through a small trainable bridge. The bridge is usually treated as plumbing, but it in fact defines the geometry of the speech-to-LLM interface, and the pretraining objective decides whether that interface provides a reusable initialization for downstream tasks. We study a transferable bridge through two complementary properties: global alignment with the text side, and local lexical manifold compatibility, where bridge embeddings remain close to the frozen LLM's input-embedding neighbourhoods. We make this property measurable with a fixed, head-free, timestamp-free diagnostic that applies to any objective, and show that next-word prediction (NWP) and sentence-level contrastive pretraining do not fully capture token-level lexical compatibility. We then introduce Local Prototype Reconstruction (LPR), a lightweight training-only regularizer that requires each aligned bridge token to be reconstructable from a small neighbourhood of frozen LLM token embeddings, with a hard single-prototype anchor as its limiting case. On multilingual ASR and speech translation, LPR improves transfer,…

    PDF
  • arXiv cs.AIOct 8

    Social Pain Disrupts Emotion-Action Brain-State Dynamics in Adolescents with Non-Suicidal Self-Injury

    Ying Xu, Xiaojun Liang, Li Zhang, Yixuan Yuan, Gan Huang +2

    Non-suicidal self-injury (NSSI) is prevalent among adolescents with depression, but the rapid brain-state dynamics linking social distress to maladaptive behavior remain unclear. We combine an experimental pain paradigm, electroencephalography (EEG) microstate analysis, and interpretable deep sequence modeling to investigate NSSI-related neurodynamics in 106 adolescents with depression, including 67 with NSSI (DN+) and 39 without NSSI (DN-), during social pain, physical pain, and resting-state conditions. A model integrating disease-specific, domain-adversarial, consistency, and contrastive learning captures higher-order dependencies in microstate sequences. Social pain yields the strongest NSSI discrimination, with 68.55% accuracy, outperforming the best baseline by 8.94% points. Model interpretation and conventional microstate analyses reveal weakened bidirectional transitions between MS3 and MS5 in DN+ adolescents during social pain. Source reconstruction associates MS3 with emotional/interoceptive processing and MS5 with action preparation, suggesting disrupted emotion-action coupling. Time-resolved analyses show greater early-to-middle action-state recruitment and later…

    PDF
  • arXiv cs.CVOct 8

    CARE: Constrained Attention Refinement for Fine-Grained Visual Classification via Teacher-Student Distillation

    Ruibo Wen, Hang Shao, Yiming Lei

    Fine-grained visual classification requires models to recognize subtle local traits while exposing the visual evidence behind their predictions. Class-specific attention pathways provide a natural basis for interpretable recognition, but their constrained prediction structure limits discriminative capacity and underuses intermediate representations from strong pretrained backbones. To address this problem, we propose CARE, a constrained attention refinement framework for interpretable fine-grained recognition via teacher-student distillation. CARE keeps the final prediction and explanation within a class-specific attention student, while introducing a training-only auxiliary query teacher that reads selected intermediate DINOv2 layers with learnable queries. The teacher fuses multi-level representations and transfers logit-standardized class-discriminative knowledge to the student. To further refine the explanation pathway, we design diversity and sparsity terms to regularize student attention heads, reducing redundancy and encouraging compact trait localization. Experiments on CUB, Oxford-IIIT Pet, Stanford Dogs, and Stanford Cars show that CARE achieves strong classification…

    PDF
  • arXiv cs.LGOct 8

    Do LLMs Learn from Rewards in Context? : Rethinking the role of reward in In-Context RL

    Minchan Kwon, Seunghee Koh, Sunghyun Baek, Minsung Bae, Junmo Kim

    LLM agents increasingly improve at inference time by accumulating experience in context rather than by updating parameters. This process is often described as in-context reinforcement learning (ICRL). Whether in-context learning (ICL) can actually play the role of RL, however, has not been tested. We study this question in its simplest form, direct ICRL, where the model conditions directly on raw trajectory-reward pairs, and ask whether the reward acts as a learning signal. Through controlled experiments on four benchmarks across six models, we find that the reward is read, but its effect is small: flipping, randomizing, or removing the reward leaves the improvement curve almost unchanged, and this holds even under meta-prompts that explicitly instruct the model to explore, exploit, or reason over rewards. Trajectories drive improvement, but not through their semantic content: shuffled or corrupted trajectories work as well as real ones. These patterns closely mirror those known in ICL, suggesting that direct ICRL is better understood as a special case of ICL than as inference-time RL. This reframing has implications for agent memory design: ICL factors such as input distribution…

    PDF
  • arXiv cs.CVOct 8

    SP-DocReader: Difference-Aware Self-Play for Precise Document OCR

    Wenjie Liao, Xiaohui Song, Liangjie Zhao, Haonan Lu

    Accurate page transcription remains difficult for vision language models under limited input and training budgets. We present SP-DocReader, a self-play framework for optical character recognition (OCR) that targets residual errors after supervised fine-tuning. Reading Discrepancy Masking aligns reference and generated model tokens through a longest common subsequence, then scores unmatched positions with their full conditioning prefixes. Focused Fidelity Loss adds direct negative log-likelihood supervision at unmatched ground-truth positions. Only the OCR module is trained, while the backbone remains frozen. We derive the combined gradient to distinguish relative score optimization from direct supervision. Compared with SFT-2, SP-DR-3 reduces Vary-600K character error rate on both backbones. On Qwen3-VL-4B, it reduces character error rate by approximately 54 percent and improves DocVQA Average Normalized Levenshtein Similarity (ANLS) by 3.7 points. These results show the value of focusing self-play training on the discrepancies that remain after supervised fine-tuning.

    PDF
  • arXiv cs.CVOct 8

    A Unified Score Matching Paradigm for Video Anomaly Detection and Anticipation

    Congqi Cao, Zhenhe Liang, Hanwen Zhang, Yifan Zhao, Qinyi Lv +2

    Video anomaly detection (VAD) is a fundamental and safety-critical task in computer vision. Recent generative approaches detect anomalies from a distributional perspective, but remain limited by local anomaly modes. Meanwhile, video anomaly anticipation (VAA), as a proactive extension beyond post-hoc detection, introduces additional challenges. In particular, the contrastive inference paradigm in VAD, which relies on ground-truth frames, is not applicable to VAA, hindering its development. To address these challenges, we propose a unified score-driven framework, termed Uni-DSM, based on denoising score matching (DSM), which models anomaly patterns through likelihood estimation and score functions over the learned data distribution. Within this unified framework, we adopt a shared noise-conditioned score transformer backbone with scene-dependent embeddings and motion-aware weighting for distribution-level modeling. Instead of introducing separate architectures, Uni-DSM unifies VAD and VAA through different inference and supervision paradigms built upon the same score-based formulation. For VAD, we instantiate an autoregressive denoising score matching (ADSM) mechanism, which…

    PDF
  • arXiv cs.LGOct 8

    When Do External Priors Help Low-Resource Structured Prediction? A Ranking-Level Alignment Study

    Qiye Lu, Jiang Ji, Liang Zhang

    Cold-start credit scoring (scarce labels, immature features, or minimal capacity) motivates injecting external priors (teacher models, LLMs, or domain experts) into training. We study Ranking Prior Alignment (RPA), a model-agnostic objective that aligns a model's ranking distribution to an external prior via a temperature-scaled KL loss with an annealed weight, $\mathcal{L} = \mathcal{L}_{\text{task}} + γ(t)\,\text{KL}(P_{\text{agent}} \| P_{\text{model}})$, and requires no external model at inference. Across tree-ensemble and multiple-instance-learning (MIL) instantiations on the public Amex dataset and an industrial merchant-credit dataset (1.5M+ merchants), we ask when such priors genuinely transfer information rather than merely act as a regularizer. Our central finding: the prior's quality $π$ (its rank correlation with the label) governs transfer only when the alignment granularity matches the granularity at which $π$ is measured. On cross-sample tree alignment (Amex), a high-$π$ teacher yields $+0.028$ AUC over the no-alignment baseline (paired bootstrap 95% CI $[0.022, 0.035]$) whereas zero-information priors (uniform, random) yield none, and $Δ$AUC tracks $π$ (XGBoost…

    PDF
  • arXiv cs.CVOct 8

    Improving Image-Based Nutrition Estimation Through Multimodal Food-Item Verification and Recovery

    Jingbo Yue, Bruce Coburn, Jinge Ma, Jui-Feng Chi, Fengqing Zhu

    Single-image nutrition estimation can fail silently when visible foods are missed. Even when a food is correctly identified, its proposed region may not support portion estimation. We propose a framework that uses multimodal large language models (MLLMs) to inventory visible foods and separately verify food identity and whether each proposed 2D region supports portion estimation. One whole-image review uses these verification results to identify unresolved gaps and omitted foods, triggering at most one targeted recovery pass. Recovered regions are re-verified without access to the recovery prompt, then reconciled into a final item set for nutrition estimation. The framework requires no task-specific fine-tuning. Matched evaluation on common valid-output samples shows that item-level grounding improves mass accuracy across all tested settings and energy accuracy relative to an adapted retrieval baseline, with item-identity precision and recall also improving, while post-recovery visual coverage is assessed separately at inference time without ground-truth annotations.

    PDF
  • arXiv cs.LGOct 8

    ActiveMedAgent: Cost-Aware Trajectory Learning for Multimodal Medical Diagnosis

    Weiwei Ma, Xiaobing Yu, Peijie Qiu, Jin Yang, Zhaoqi An +3

    Clinical diagnosis is inherently sequential: clinicians escalate from cheap to costly tests only when additional evidence is expected to resolve diagnostic uncertainty. We present ActiveMedAgent, a framework that brings this cost-aware sequential logic to multimodal medical AI. Given a frozen, API-accessed vision-language model, ActiveMedAgent tracks probability distributions over candidate diagnoses and scores each acquisition by its per-step diagnostic utility minus cost. A lightweight MLP controller is then trained offline on these scored trajectories, learning when to request additional evidence and when to commit. Across three commonly used benchmarks, trajectory-based policy learning consistently outperforms both unguided acquisition and full-modality baselines. Notably, we identify an information overload effect. In 175 cases, the agent produces a correct diagnosis with fewer channels while the full-modality baseline fails, showing that learning what to omit can be as important as learning what to acquire.

    PDF
  • arXiv stat.MLOct 8

    Accelerating Non-Smooth and Heavy-Tailed Sampling

    Pervez Ali, Xiaoyu Wang, Yingli Wang, Lingjiong Zhu

    Anchored Langevin dynamics (ALD) is useful for non-smooth sampling where the density of the target distribution is possibly non-differentiable and heavy-tailed; reflected anchored Langevin dynamics (RALD) can sample possibly non-differentiable target density on a constrained domain. In this paper, we propose and study non-reversible anchored Langevin dynamics (NALD) for sampling possibly non-differentiable and heavy-tailed target density in the Euclidean space and the non-reversible reflected anchored Langevin dynamics (NRALD) for sampling possibly non-differentiable target density in the constrained space. Our construction adds a circulation drift generated by a possibly state-dependent divergence-free skew-symmetric matrix field and a stream potential. It preserves the target distribution without requiring derivatives of target density, admits a random-time-change representation, and applies both on the whole Euclidean space and on bounded domains with normal reflection. By breaking reversibility, we show that NALD and NRALD can converge to their target distributions faster than their reversible counterparts via finite-time non-asymptotic convergence analysis, a large…

    PDF
  • arXiv cs.CVOct 8

    IntrinSync: Joint Intrinsic Decomposition and Reciprocal Rendering

    Zheng Gu, Rui Huang, Xilu Zhang, Jingbo Zhang, Min Lu +4

    Inverse rendering decomposes an image into intrinsic properties such as appearance, illumination, geometry, and material, yet these properties are inherently interdependent. A reliable decomposition should produce intrinsic maps that are not only individually plausible, but also mutually compatible in explaining the image. However, existing methods either model intrinsic channels in isolation or treat inverse and forward rendering as separate processes, leaving the interdependence underexploited. In this paper, we introduce IntrinSync, a unified framework that captures this interdependence through joint-channel modeling and reciprocal inverse-forward rendering. At the channel level, we jointly decompose an input RGB into albedo, shading, surface normal, roughness, and metallic maps through a 1-to-N mapping, enabling information exchange across channels throughout generation. At the process level, we establish inverse-forward reciprocity through a dual cycle-consistent objective that aligns corresponding predictions across a closed loop. Experiments on three datasets demonstrate that our method achieves competitive intrinsic estimation and forward rendering performance, improving…

    PDF
  • arXiv cs.CLOct 8

    The "10th Juror": Open-Set Standpoint Screening for Bureaucratic Bias Detection

    Yuchen Miao, Zijun Wang, Chang Han, Yurui Shi, Mingtai Zhang +1

    Presupposing the boundaries of bias is itself a form of bias. We study closed-loop bias governance for Dutch government documents, where a system must detect biased language, ground decisions in legal and contextual evidence, rewrite problematic sentences when intervention is warranted, and verify that the rewrite mitigates harm without distorting meaning. Existing methods face three challenges: (i) discriminative classifiers capture surface regularities but lack normative grounding; (ii) zero-shot LLMs often adopt generic viewpoints and over-flag ambiguous administrative language; and (iii) fixed taxonomies inherit the Closed-World Assumption, missing emerging local targets. We propose MARS-Gov, a standpoint-aware multi-agent framework that combines legal retrieval, open-set target screening, specialized jurors, conservative routing, and rewrite verification. When screening finds an uncovered group, MARS-Gov instantiates a dynamic "10th juror" to deliberate outside the fixed panel. On DGDB, MARS-Gov sets a new SOTA with 0.880 F1, outperforming the strongest zero-shot LLM detector by 20.2 points (29.8% relative) and the best supervised Dutch encoder by 6.8 points, while reducing…

    PDF
  • arXiv cs.CLOct 8

    Can a System-One LLM Perform Knowledge Tracing When Few or No Learners Are Logged?

    Unggi Lee, Haeun Park

    Knowledge tracing (KT) models need many logged learners, so a new course or platform starts without a usable model. In LLM-based KT the LLM generates the answer, which we call System-Two; it is either fine-tuned on the target data or reasons and votes over ten samples, which is slow and gives coarse probabilities. We ask whether an off-the-shelf System-One LLM, which returns a probability for a typed question directly in a single pass, can perform KT when few or no learners are logged. On seven datasets, Jev without any data from the target platform reaches a mean AUC of .706, above the best of 28 deep KT models trained on 8 learners (.689) and above System-Two Thinking-KT on all seven datasets (.650) at about 1/100 of its API cost. Adding examples and a similar-learner statistic from the logged learners (JevKT) raises this to .722; JevKT stays significantly ahead of deep KT up to 16 learners and ahead on average up to 64, and supervised KT catches up between 64 and 128 learners. Among the readers we tested, the gain is specific to Jev, since three other LLMs queried with the byte-identical typed request through the official System-One adapter fall below it on all seven datasets,…

    PDF
  • arXiv cs.LGOct 8

    QUILT: Rethinking Sparse-Attention Prefill through Shared Query Execution

    Zhenduo Zhao, Qihui Zhou, Mingcong Song, Zhiyi Chen, Chuangguan Ye +5

    Sparse attention reduces the cost of long-context attention, but existing kernels typically process queries independently, repeatedly loading and dequantizing KV entries shared across queries. We observe substantial overlap in the KV entries selected by neighboring queries, creating opportunities for cross-query reuse. We present QUILT, a workload-aware sparse-attention execution mechanism that jointly processes neighboring queries and reuses shared KV entries to reduce redundant memory traffic and computation. QUILT introduces Shift-and-Compare Set Decomposition (SCSD), which transforms irregular set operations into regular data-parallel primitives suitable for modern accelerators, and pipelines SCSD with attention computation to hide its overhead. Cascaded sharing captures reuse hierarchically at multiple granularities. A tile-aware execution strategy balances sharing granularity with hardware tile utilization and selectively removes low-importance query-specific tails to eliminate underutilized tiles. We evaluate QUILT on LongBench using GLM-5.3 and DeepSeek-3.2 under both tensor and sequence parallelism. Compared with the state-of-the-art sparse-attention kernel, QUILT…

    PDF
  • arXiv cs.LGOct 8

    Machine Learning Optimization for Enhanced OS Fingerprinting

    Jae Sung Kim, Spencer Ekeroth, Jeremy Neale

    Operating System (OS) Fingerprinting is a technique that can be used to identify a network's operating systems by evaluating network traffic in the form of TCP/IP packets. This research will explore the effectiveness of passively identifying operating systems on the CIC-IDS2017 dataset, a collection of over 47 gigabytes of pcap files with their corresponding operating systems. This research also proposes a new command line interface, OsirisML, which uses nPrint to preprocess the data into tabular data and XGBoost to apply ML to the data to generate, retrain, and test ML models. When packets are split randomly between training and testing, OsirisML models reach an accuracy of 97.66% on a down-sampled subset of the Friday capture and 84.69% on the entire capture. On the entire Monday capture, which contains no attacks, OsirisML reaches an accuracy of 73.83% and an F-1 score of 79.38%.

    PDF
  • arXiv cs.CLOct 8

    SFT-as-Context Mitigates Forgetting in Supervised Fine-Tuning

    Kenan Tang, Andong Hua, Chengxuan Qian, Saket Tiwari, Yao Qin

    Supervised fine-tuning (SFT) equips large language models (LLMs) with specialized capabilities, but often comes at the cost of forgetting the general capabilities of their parent models (i.e., the pretrained models before fine-tuning). This trade-off is especially limiting for queries that require both specialized and general capabilities. We introduce SFT-as-context, a training-free method in which the parent model uses the SFT model's response as context to answer the query. This allows the parent model to acquire fine-tuned capabilities from the SFT response through in-context learning while preserving its own general capabilities. Across 19 parent-SFT model pairs and 11 benchmarks, SFT-as-context remains close to the SFT models on fine-tuned capabilities, with gaps of only 2.2 and 2.1 percentage points on AIME 2024 and LiveCodeBench and 2.0 macro MAE on NutriBench-English, while staying within 2.2 percentage points of the parent models on general capabilities on average. Remarkably, it can solve queries requiring both fine-tuned and general capabilities, even when neither the parent nor SFT model succeeds alone. This approach also extends beyond parent-SFT pairs: responses…

    PDF
  • arXiv cs.AIOct 8

    GameCommBench: A Unified Benchmark and Type-Aware Evaluation for AI-Generated Game Commentary

    Qirui Zheng, Zhengteng Lin, Yunyi Xiao, Junhao Li, Keyuan Cheng +5

    Game commentary is an open-ended generation task requiring multimodal perception, strategic reasoning, and contextual knowledge. Existing AI-Generated Game Commentary (AI-GGC) studies remain fragmented across games, modalities, and evaluation protocols, while overlap-based or holistic evaluators fail to capture the functional heterogeneity of commentary. We introduce \textsc{GameCommBench}, a unified benchmark spanning board games, sports, and esports, with commentary aligned to heterogeneous game contexts and annotated by commentary type. We further propose Type-Aware Commentary Evaluation (TACE), a structured framework for evaluating different types of commentary. We then validate TACE for reliability and human agreement, and use it to benchmark representative AI commentators. Results reveal non-uniform capability profiles, with live observation and strategic analysis emerging as major bottlenecks. Together, \textsc{GameCommBench} and TACE provide a diagnostic foundation for comparable and interpretable AI-GGC evaluation.

    PDF
  • arXiv cs.AIOct 8

    Balancing Reference Guidance and Free Generation in Trajectory Rollouts for Reasoning RL

    Hanyu Wang, Nakul Agarwal, Hossein Nourkhiz Mahjoub, Ehsan Moradi Pari, Makoto Fukushima +2

    A verified reference solution provides a correct trajectory for training a reasoning model. Alternatively, a prefix of the reference can guide the model in generating a trajectory of its own. How much reference guidance should we provide? We study this question through prefix continuation, where the model continues from a reference prefix and keeps the resulting trajectory if it passes verification, falling back to the reference otherwise. Since both procedures produce correct trajectories, we compare their distributions with the ideal distribution, the model's own distribution conditioned on successful verification. For one continuation, we derive the KL divergence in closed form, which, up to a bounded term, decreases with the product of the probability of generating a different correct trajectory and the reference surprisal, the negative log probability of the reference suffix given the prefix. Since a longer prefix tends to raise the former but lowers the latter, continuation success alone does not determine the preferred amount of guidance. From this analysis, we learn a prefix selector shared across training questions from continuation outcomes, without estimating success…

    PDF
  • arXiv cs.AIOct 8

    When Interfaces Speak: Data-Aware Generative UI Harness for Active Interaction

    Xiaolong Li, Xiaohan Xu, Jinyang Li, Xinnuo Xu, Ge Qu +3

    Most human-agent interaction today remains text-based. Natural language can impose cognitive overload, ambiguity, information chaos, and slow input for complex tasks; ephemeral generative UIs can present structured information and guide users toward task completion. We propose GenUI-Harness, a multi-agent harness pairing a Tool Agent for information retrieval and task execution with a GUI Coder Agent that identifies ambiguities and generates front-end code for structured interfaces. Training the coder with reinforcement learning is challenging: verifiable rewards for interactive UI generation require costly execution, while LLM-as-a-Judge rewards are prone to reward hacking. We address the first challenge with Dynamic UX, a lightweight package for dynamic interaction and reward collection in a single sandbox, and the second with Reward Auditor, a meta-reward mechanism that monitors reward distributions and distills diagnostic patterns into a shared rubric and scoring specification. We introduce UI-TAU Bench, a benchmark for active human-agent interaction through generated UI code, built on 10 real-world domain databases constructed from public data sources and based on Tau-Bench…

    PDF
  • arXiv cs.CLOct 8

    CoTS: Competitive Token Selection for Latent Reasoning in Large Language Models

    Jie Yu, Yuxi Sun, Zhaoxu Li, Chunxiao Liu, Xudong Jiang

    Recent latent reasoning methods move beyond discrete Chain-of-Thought (CoT) by replacing sampled tokens with continuous soft embeddings or hidden states, enabling richer intermediate representations and preserving alternative reasoning paths. However, the need for latent reasoning varies across decoding steps: confident predictions favor discrete commitment, whereas ambiguous steps benefit from preserving multiple plausible alternatives. Existing approaches typically handle this variation through either global soft activation or hard switching based on uncertainty thresholds, treating latent intervention as an all-or-nothing decision rather than continuously adapting its strength. To address this limitation, we propose Competitive Token Selection (CoTS), a training-free framework that reframes latent intervention from binary mode selection into continuous allocation at the token level. CoTS uses the probability competition between the two most likely candidates to dynamically balance discrete commitment and soft exploration at each reasoning step. The second candidate represents the strongest alternative to the model's preferred continuation, so the separation between the two…

    PDF
Load more