Stop When Enough: Adaptive Early-Stopping for Chain-of-Thought Reasoning

/in /by

Chain-of-Thought (CoT) reasoning has driven recent gains of large language models (LLMs) on reasoning-intensive tasks by externalizing intermediate steps. However, excessive or redundant reasoning — so-called overthinking— can increase inference costs and lead LLMs toward incorrect conclusions. In this paper, we present REFRAIN (REFlective-Redundancy for Adaptive INference), a training-free framework that adaptively determines when to stop reasoning to mitigate overthinking. REFRAIN integrates a two-stage stop discriminator to identify reflective yet redundant reasoning and a sliding-window Upper Confidence Bound (SW-UCB) multi-armed bandit controller to dynamically adjust stopping thresholds according to problem difficulty without supervision or fine-tuning. Across four representative benchmarks and two model families, REFRAIN re-duces token usage by 20-55% while maintaining or improving accuracy compared to standard CoT prompting. Extensive ablation and robustness analyses demonstrate its stability across models, scorers, and prompt variations. In summary, our findings highlight when-to-stop as a new and practical axis of test-time scaling — enabling models to reason not just more, but just enough.

Read more

Representation Interventions Enable Lifelong Knowledge Memory Control in LLMs

/in /by

Large language models (LLMs) often produce incorrect or outdated content after being employed. Efficient and accurate knowledge up-dates without costly retraining are a major challenge. This problem is particularly challenging in lifelong settings, where complex, unstructured knowledge must coexist without interference. We introduce RILKE (Representation Intervention for Lifelong KnowledgE Control), a robust and scalable method that treats knowledge control as interventions within the model’s representation space. Leveraging representation-space expressiveness, we identify two key properties enabling RILKE to achieve fine-grained control over complex, un-structured knowledge while maintaining general utility with frozen base weights. During training, RILKE learns paraphrase-robust and edit-localized modules that limit each update to a low-dimensional subspace to minimize cross-edit interference. At inference, a query-adaptive router selects the appropriate module to guide the model’s generation. Across LLaMA and Qwen models, RILKE scales effectively to large-scale benchmarks, demonstrating high edit success and strong paraphrase generalization while preserving general utility with modest memory overhead. These results show RILKE is an effective and scalable solution for lifelong knowledge control in LLMs.

Read more

Uncertainty-Aware Test-Time Search for Optimization Problem Solving

/in /by

Automatically solving optimization problems from natural language descriptions with both efficiency and reliability is highly desirable but remains challenging. Language model hallucinations and the limited availability of labeled datasets often result in misaligned formulations, code errors, and feasibility failures. We pro-pose UMCTS, an Uncertainty-aware Monte Carlo Tree Search framework that combines the language understanding capability of large language models with the reliability of well-established solvers. UMCTS structures the solution process into four stages: global instruction, assumptions, mathematical formulation, and solver code generation. It employs Monte Carlo Tree Search with semantic-equivalence pruning, prior-guided exploration, and solver-based feasibility checks. An LLM judge provides numerical reward signals, qualtative error information, and uncertainty estimates. These signals are backpropagated to guide the search and flag unreliable out-puts. Across six public benchmarks, UMCTS achieves state-of-the-art solution accuracy and improves efficiency by reducing token usage.

Read more

Mind the Gap in Cultural Alignment: Task-Aware Culture Management for Large Language Models

/in /by

Large language models (LLMs) are increasingly deployed in culturally sensitive real-world tasks. However, existing cultural alignment ap-roaches fail to align LLMs’ broad cultural values with the specific goals of downstream tasks and suffer from cross-culture interference. We propose CultureManager, a novel pipeline for task-specific cultural alignment. CultureManager synthesizes task-aware cultural data in line with target task formats, grounded in culturally relevant web search results. To prevent conflicts between cultural norms, it manages multi-culture knowledge learned in separate adapters with a culture router that selects the appropriate one to apply. Experiments across five national cultures and ten culture-sensitive tasks show consistent improvements over prompt-based and fine-tuning baselines. Our results demonstrate the necessity of task adaptation and modular culture management for effective cultural alignment.

Read more

LOKA: Conflict-Aware LLM Knowledge Update with Adaptive Knowledge Memory

/in /by

Large Language Models (LLMs) have achieved remarkable success in natural language processing by encoding extensive knowledge, but their utility relies on timely updates as human knowledge keeps evolving. In this paper, we investigate the problem of LLM knowledge updates, which requires simultaneously unlearning un-wanted information and learning new knowledge. Existing approaches that tackle unlearning and learning separately encounter task conflicts and knowledge management issues when applied to comprehensive knowledge updates. In this paper, we validate our findings with theoretical analysis and empirical evidence, and propose LOKA, a conflict-aware framework for Large language mOdel Knowledge updAtes. During training, LOKA introduces an adaptive knowledge memory approach in which updated knowledge is allocated across multiple memory units. During inference, LOKA retrieves the most relevant memory unit from the knowledge memory and integrates it with the original LLM to apply updated knowledge, while a learning-based router controls the activation of the knowledge memory to improve knowledge utilization. Extensive experiments demonstrate the efficacy of LOKA in achieving accurate, flexible, and conflict-aware knowledge updates.

Read more

Subgroup Discovery with the Cox Model

/in /by

We study the problem of subgroup discovery for survival analysis, where the goal is to find an interpretable subset of the data on which a Cox model is highly accurate. We examine why existing quality functions are insufficient for this problem and introduce two technical innovations: the expected prediction entropy (EPE), a novel metric for evaluating survival models that predict hazard functions, and the conditional rank statistics (CRS), which quantifies individual point deviation from a subgroup’s survival time distribution. We study the EPE and CRS theoretically and show they address problems with existing metrics. We then introduce eight algorithms for Cox subgroup discovery. Our main algorithm is based on the DDGroup framework of Izzo et al. (2023) and leverages both the EPE and CRS, allowing theoretical correctness guarantees in well-specified settings. Empirical evaluation on synthetic and real data confirms our theory, showing our methods recover ground-truth subgroups in well-specified cases and achieve better model fit than naively fitting the Cox model to the entire dataset. A case study on NASA jet engine simulation data demonstrates that discovered subgroups uncover known nonlinearities in the data and suggest design choices mirrored in practice.

Read more

Logical Guidance for the Exact Composition of Diffusion Models

/in /by

We propose LOGDIFF (Logical Guidance for the Exact Composition of Diffusion Models), a guidance framework for diffusion models that enables principled constrained generation with complex logical expressions at inference time. We study when exact score-based guidance for complex logical formulas can be obtained from guidance signals associated with atomic attributes and constraints.First, we derive an exact Boolean calculus that provides a sufficient condition for exact logical guidance. Specifically, if a formula admits a circuit representation in which conjunctions combine conditionally independent subformulas and disjunctions combine subformulas that are either conditionally independent or mutually exclusive, exact logical guidance is achievable. In this case, the guidance signal can be computed exactly from atomic scores and posterior probabilities using an efficient recursive algorithm. Moreover, we show that, for commonly encountered classes of distributions, any desired Boolean formula is compilable into such a circuit representation.Second, by combining atomic guidance scores with posterior probability estimates, we introduce a hybrid guidance approach that bridges classifier guidance and classifier-free guidance, applicable to both compositional logical guidance and standard conditional generation.We demonstrate the effectiveness of our framework on multiple image and protein structure generation tasks.

Read more

Rapid State-of-Polarization Change-Point Detection via Minimum Lossy Coding Length

/in /by

We propose a rapid state-of-polarization disturbance change-point detection and localization method using lossy coding length of kernel features. Field experiments validateaccurate detection of subtle changes and precise localization, enabling reliable fiber identification and monitoring.

Read more

Multi-Fiber/Multi-DAS Array System with Clock-Free Synchronization

/in /by

We demonstrate distributed fiber-optic acoustic sensing that employs a mechanical synchronizer to achieve sample-level synchronization of signals across independent sensors.This enables phase-noise suppression and bandwidth extension in standard multi-fiber optical cables without hardware modification.

Read more

Mobile Orbital Domain-based Hierarchical Routing with Joint Path and Gateway Selection

/in /by

We propose to jointly optimize path and gateway selection to achieve load balancing for mobile orbital domain-based hierarchical routing in satellite networks. The load at bottleneck gateway satellites can be reduced by 22% on average.

Read more