This paper presents a novel approach to avoiding jackknifing and mutual collisions in Heavy Articulated Vehicles (HAVs) by leveraging decentralized swarm intelligence. In contrast to typical swarm robotics research, our robots are elongated and exhibit complex kinematics, introducing unique challenges. Despite its relevance to real-world applications such as logistics automation, remote mining, airport baggage transport, and agricultural operations, this problem has not been addressed in the existing literature. To tackle this new class of swarm robotics problems, we propose a purely reaction-based, decentralized swarm intelligence strategy tailored to automate elongated, articulated vehicles. The method presented in this paper prioritizes jackknifing avoidance and establishes a foundation for mutual collision avoidance. We validate our approach through extensive simulation experiments and provide a comprehensive analysis of its performance. For the experiments with a single HAV, we observe that for 99.8% jackknifing was successfully avoided and that 86.7% and 83.4% reach their first and second goals, respectively. With two HAVs interacting, we observe 98.9%, 79.4%, and 65.1%, respectively, while 99.7% of the HAVs do not experience mutual collisions.

Contextual predictability shapes how we choose and encode words in production. The effects of a word's predictability given preceding or past context are generally well-understood in both production and comprehension, but studies of naturalistic production have also revealed a poorly-understood yet robust backward predictability effect of a word given only its future context, which may be linked to future planning. Across two studies of naturalistic speech, we revisit backward predictability using improved operationalizations, introducing a conceptually motivated information-theoretic measure that quantifies the information shared between a word and future context under the constraints imposed by the past context. Study 1 shows that this measure produces effects qualitatively similar to backward predictability while explaining unique variance in phonetic reduction. Study 2 examines substitution errors within a generative framework that models lexical, contextual, and communicative influences on word choice to predict the identity of the word that surfaces as an error. Within this framework, we find that past-conditioned predictability increases error likelihood, whereas future-conditioned predictability reduces it. Further, our proposed measure emerges as the strongest contextual predictor of error identity, subsuming backward predictability. Analysis of error types further reveals graded trade offs in how speakers prioritize form-, meaning-, and context-based information during lexical planning. Together, these findings illuminate how past and future context shape word choice and encoding, linking contextual predictability to mechanisms of incremental planning in sentence production.

LLMs can solve complex tasks by generating long, multi-step reasoning chains. Test-time scaling (TTS) can further improve performance by sampling multiple variants of intermediate reasoning steps, verifying their correctness, and selecting the best steps for continuation. However, existing verification approaches, such as Process Reward Models (PRMs), are computationally expensive and require large-scale human or model-generated annotations. We propose a lightweight alternative for step-level reasoning verification based on probing the internal states of LLMs. We train a transformer-based probe that uses the internal states of a frozen LLM to estimate the credibility of its reasoning steps during generation. Annotation can be provided either by a larger LLM (e.g., DeepSeek-R1) or in a self-supervised manner by the original model itself. The probes are lightweight, containing fewer than 10M parameters. Across multiple domains, including mathematics, planning, and general knowledge question answering, our probes match or exceed the performance of PRMs that are up to 810x larger. These results suggest that LLM internal states encode confidence in their reasoning processes and can serve as reliable signals for step verification, offering a promising path toward scalable, generalizable TTS and more introspective LLMs.

A common approach to mechanistic interpretability is to causally manipulate model representations via targeted interventions in order to understand what those representations encode. Here we ask whether such interventions create out-of-distribution (divergent) representations, and whether this raises concerns about how faithful their resulting explanations are to the target model in its natural state. First, we demonstrate theoretically and empirically that common causal intervention techniques often do shift internal representations away from the natural distribution of the target model. Then, we provide a theoretical analysis of two cases of such divergences: "harmless" divergences that occur in the behavioral null-space of the layer(s) of interest, and "pernicious" divergences that activate hidden network pathways and cause dormant behavioral changes. Finally, in an effort to mitigate the pernicious cases, we apply and modify the Counterfactual Latent (CL) loss from Grant (2025) allowing representations from causal interventions to remain closer to the natural distribution, reducing the likelihood of harmful divergences while preserving the interpretive power of the interventions. Together, these results highlight a path towards more reliable interpretability methods.

Recently, Mamba-based super-resolution (SR) methods have demonstrated the ability to capture global receptive fields with linear complexity, addressing the quadratic computational cost of Transformer-based SR approaches. However, existing Mamba-based methods lack fine-grained transitions across different modeling scales, which limits the efficiency of feature representation. In this paper, we propose T-PMambaSR, a lightweight SR framework that integrates window-based self-attention with Progressive Mamba. By enabling interactions among receptive fields of different scales, our method establishes a fine-grained modeling paradigm that progressively enhances feature representation without introducing additional computational cost. Furthermore, we introduce an Adaptive High-Frequency Refinement Module (AHFRM) to recover high-frequency details lost during Transformer and Mamba processing. Extensive experiments demonstrate that T-PMambaSR progressively enhances the model's receptive field and expressiveness, yielding better performance than recent Transformer- or Mamba-based methods while incurring lower computational cost.

We theoretically investigate whether the Query, Key, Value weight triplet can be reduced in encoder-only and decoder-only transformers. Under mild assumptions, we prove that one of the Query, Key or Value weights are redundant and can be replaced with the identity matrix, reducing attention parameters by 25\%. If applied to the Query or Key weights, this also simplifies optimization: attention logits depend on a single learned weight matrix rather than on a product of two. Validating the Query weight removal on decoder-only GPT-style small models trained from scratch, we find that reduced models match baseline performance despite fewer parameters, and outperform baselines when saved parameters are reallocated. Our analysis has also led us to a structural expressivity boundary: in the mathematically tractable ReLU setting, skip connections push MLPs into a generically disjoint function class at fixed width. These findings motivate investigation across modalities and at scale, where the observed stability and efficiency gains may prove most consequential.

Retrieval-augmented generation (RAG) remains brittle on multi-step questions and heterogeneous evidence sources, trading accuracy against latency and token/tool budgets. This paper introduces RELOOP, a structure aware framework using Hierarchical Sequence (HSEQ) that (i) linearize documents, tables, and knowledge graphs into a reversible hierarchical sequence with lightweight structural tags, and (ii) perform structure-aware iteration to collect just-enough evidence before answer synthesis. A Head Agent provides guidance that leads retrieval, while an Iteration Agent selects and expands HSeq via structure-respecting actions (e.g., parent/child hops, table row/column neighbors, KG relations); Finally the head agent composes canonicalized evidence to genearte the final answer, with an optional refinement loop to resolve detected contradictions. Experiments on HotpotQA (text), HybridQA/TAT-QA (table+text), and MetaQA (KG) show consistent EM/F1 gains over strong single-pass, multi-hop, and agentic RAG baselines with high efficiency. Besides, RELOOP exhibits three key advantages: (1) a format-agnostic unification that enables a single policy to operate across text, tables, and KGs without per-dataset specialization; (2) \textbf{guided, budget-aware iteration} that reduces unnecessary hops, tool calls, and tokens while preserving accuracy; and (3) evidence canonicalization for reliable QA, improving answers consistency and auditability.

Speculative decoding is widely used in accelerating large language model (LLM) inference. In this work, we focus on the online draft model selection problem in speculative decoding. We design an algorithm that provably competes with the best draft model in hindsight for each query in terms of either the token acceptance probability or expected acceptance length. In particular, we show that we can accurately evaluate all draft models, instead of only the chosen model without incurring additional queries to the target model, which allows us to improve exponentially over the existing bandit-based approach as the number of draft models increases. Our approach is generically applicable with any speculative decoding methods (single draft, multi-drafts and draft-trees). Moreover, we design system-efficient versions of online learners and demonstrate that the overhead in computation and latency can be substantially reduced. We conduct extensive experiments on open-source LLMs and diverse datasets, demonstrating that our methods substantially outperform the state-of-the-art EAGLE3 and the BanditSpec baseline in a variety of domains where specialized domain-expert drafters are available, especially when long reasoning chains are required.

Large language models often display undesirable behaviors embedded in their internal representations, undermining fairness, inconsistency drift, amplification of harmful content, and the propagation of unwanted patterns during extended dialogue and conversations. Although training-time or data-centric methods attempt to reduce these effects, they are computationally expensive, irreversible once deployed, and slow to adapt to new conversational contexts. Pruning-based methods provide a flexible and transparent way to reduce bias by adjusting the neurons responsible for certain behaviors. However, most existing approaches are static; once a neuron is removed, the model loses the ability to adapt when the conversation or context changes. To address this, we propose a dynamic, reversible, pruning-based framework that detects context-aware neuron activations and applies adaptive masking to modulate their influence during generation. Our inference-time solution provides fine-grained, memory-aware mitigation with knowledge-preserved, more coherent behavior across multilingual single- and multi-turn dialogues, enabling dynamic fairness control in real-world conversational AI.

The performance of Latent Diffusion Models (LDMs) is critically dependent on the quality of their visual tokenizers. While recent works have explored incorporating Vision Foundation Models (VFMs) into the tokenizers training via distillation, we empirically find this approach inevitably weakens the robustness of learnt representation from original VFM. In this paper, we bypass the distillation by proposing a more direct approach by leveraging the frozen VFM for the LDMs tokenizer, named VFM Variational Autoencoder (VFM-VAE).To fully exploit the potential to leverage frozen VFM for the LDMs tokenizer, we design a new decoder to reconstruct realistic images from the semantic-rich representation of VFM. With the proposed VFM-VAE, we conduct a systematic study on how the representation from different tokenizers impact the representation learning process throughout diffusion training, enabling synergistic benefits of dual-side alignment on both tokenizers and diffusion models. Our effort in tokenizer design and training strategy lead to superior performance and efficiency: our system reaches a gFID (w/o CFG) of 2.22 in merely 80 epochs (a 10$\times$ speedup over prior tokenizers). With continued training to 640 epochs, it further attains a gFID (w/o CFG) of 1.62. These results offer solid evidence for the substantial potential of VFMs to serve as visual tokenizers to accelerate the LDM training progress.

Vision Transformers (ViTs) partition input images into uniformly sized patches regardless of their content, resulting in long input sequence lengths for high-resolution images. We present Adaptive Patch Transformers (APT), which addresses this by using multiple different patch sizes within the same image. APT reduces the total number of input tokens by allocating larger patch sizes in more homogeneous areas and smaller patches in more complex ones. APT achieves a drastic speedup in ViT inference and training, increasing throughput by 40% on ViT-L and 50% on ViT-H while maintaining downstream performance, and can be applied to a previously fine-tuned ViT, converging in as little as 1 epoch. It also significantly reduces training and inference time without loss of performance in high-resolution dense visual tasks, achieving up to 30\% faster training and inference in visual QA, object detection, and semantic segmentation.

Large Language Models (LLMs) are increasingly applied to creative domains, yet their performance in classical Chinese poetry generation and evaluation remains poorly understood. We propose a three-step evaluation framework that combines computational metrics, LLM-as-a-judge assessment, and human expert validation. Using this framework, we evaluate six state-of-the-art LLMs across multiple dimensions of poetic quality, including themes, emotions, imagery, form, and style, in the context of Tang poetry generation. Our analysis reveals a critical "echo chamber" effect: LLMs systematically overrate machine-generated poems that mimic statistical patterns yet fail strict prosodic rules, diverging significantly from human expert judgments. These findings underscore the limitations of using LLMs as standalone evaluators for culturally complex tasks, highlighting the necessity of hybrid human-model validation frameworks.

Real-world settings where language models (LMs) are deployed -- in domains spanning healthcare, finance, and other forms of knowledge work -- require models to grapple with incomplete information and reason under uncertainty. Yet most LM evaluations focus on problems with well-defined answers and success criteria. This gap exists in part because natural problems involving uncertainty are difficult to construct: given that LMs have access to most of the same knowledge as humans, it is non-trivial to design questions for which LMs will struggle to produce correct answers, but which humans can answer reliably. As a result, LM performance on reasoning under uncertainty remains poorly characterized. To address this gap, we introduce OpenEstimate, an extensible, multi-domain benchmark for evaluating LMs on numerical estimation tasks that require models to synthesize significant amounts of background information and express predictions as probabilistic priors. We assess these priors for accuracy and calibration, quantifying their usefulness relative to samples from the true distribution of interest. Across six frontier LMs, we find that LM-elicited priors are often inaccurate and overconfident. Performance improves modestly depending on how uncertainty is elicited from the model, but is largely unaffected by changes in sampling strategy, reasoning effort, or prompt design. The OpenEstimate benchmark thus offers a challenging evaluation for frontier LMs and a platform for developing models that are better at probabilistic estimation and reasoning under uncertainty.

Hate speech remains prevalent in human society and continues to evolve in its forms and expressions. Modern advancements in internet and online anonymity accelerate its rapid spread and complicate its detection. However, hate speech datasets exhibit diverse characteristics primarily because they are constructed from different sources and platforms, each reflecting different linguistic styles and social contexts. Despite this diversity, prior studies on hate speech detection often rely on fixed methodologies without adapting to data-specific features. We introduce RV-HATE, a detection framework designed to account for the dataset-specific characteristics of each hate speech dataset. RV-HATE consists of multiple specialized modules, where each module focuses on distinct linguistic or contextual features of hate speech. The framework employs reinforcement learning to optimize weights that determine the contribution of each module for a given dataset. A voting mechanism then aggregates the module outputs to produce the final decision. RV-HATE offers two primary advantages: (1)~it improves detection accuracy by tailoring the detection process to dataset-specific attributes, and (2)~it also provides interpretable insights into the distinctive features of each dataset. Consequently, our approach effectively addresses implicit hate speech and achieves superior performance compared to conventional static methods. Our code is available at https://github.com/leeyejin1231/RV-HATE.

Generating synthetic CT (sCT) from MRI or CBCT plays a crucial role in enabling MRI-only and CBCT-based adaptive radiotherapy, improving treatment precision while reducing patient radiation exposure. To address this task, we adopt a fully 3D Flow Matching (FM) framework, motivated by recent work demonstrating FM's efficiency in producing high-quality images. In our approach, a Gaussian noise volume is transformed into an sCT image by integrating a learned FM velocity field, conditioned on features extracted from the input MRI or CBCT using a lightweight 3D encoder. We evaluated the method on the SynthRAD2025 Challenge benchmark, training separate models for MRI to sCT and CBCT to sCT across three anatomical regions: abdomen, head and neck, and thorax. Validation and testing were performed through the challenge submission system. The results indicate that the method accurately reconstructs global anatomical structures; however, preservation of fine details was limited, primarily due to the relatively low training resolution imposed by memory and runtime constraints. Future work will explore patch-based training and latent-space flow models to improve resolution and local structural fidelity.

Large Language Models (LLMs) demonstrate their reasoning ability through chain-of-thought (CoT) generation. However, LLM's autoregressive decoding may limit the ability to revisit and refine earlier tokens in a holistic manner, which can also lead to inefficient exploration for diverse solutions. In this paper, we propose LaDiR} (Latent Diffusion Reasoner), a novel reasoning framework that unifies the expressiveness of continuous latent representation with the iterative refinement capabilities of latent diffusion models for an existing LLM. We first construct a structured latent reasoning space using a Variational Autoencoder (VAE) that encodes text reasoning steps into blocks of thought tokens, preserving semantic information and interpretability while offering compact but expressive representations. Subsequently, we utilize a latent diffusion model that learns to denoise a block of latent thought tokens with a blockwise bidirectional attention mask, enabling longer horizon and iterative refinement with adaptive test-time compute. This design, combined with explicit diversity guidance during diffusion inference, enables the generation of multiple diverse reasoning trajectories that explore distinct regions of the latent space, rather than producing repetitive solutions as often occurs in standard autoregressive sampling. We conduct evaluations on a suite of mathematical reasoning, code generation and puzzle planning benchmarks. Empirical results show that LaDiR consistently improves accuracy, diversity, and interpretability over existing autoregressive, diffusion-based, and latent reasoning methods, revealing a new paradigm for text reasoning with latent diffusion.

Active learning (AL) is a principled strategy to reduce annotation cost in data-hungry deep learning. However, existing AL algorithms focus almost exclusively on unimodal data, overlooking the substantial annotation burden in multimodal learning. We introduce the first framework for multimodal active learning with unaligned data, where the learner must actively acquire cross-modal alignments rather than labels on pre-aligned pairs. This setting captures the practical bottleneck in modern multimodal pipelines, where unimodal features are easy to obtain but high-quality alignment is costly. We develop a new algorithm that combines uncertainty and diversity principles in a modality-aware design, achieves linear-time acquisition, and applies seamlessly to both pool-based and streaming-based settings. Extensive experiments on benchmark datasets demonstrate that our approach consistently reduces multimodal annotation cost while preserving performance; for instance, on the ColorSwap dataset it cuts annotation requirements by up to 40% without loss in accuracy.

Recent large language models (LLMs) have shown strong reasoning capabilities. However, a critical question remains: do these models possess genuine strategic reasoning, or do they primarily excel at pattern recognition? To address this, we present ChessArena, a chess-based testbed for evaluating LLMs. Chess demands strategic reasoning, precise rule adherence, and the ability to track complex game states. ChessArena is a competitive framework where LLMs play against each other under four play modes. We evaluate 13 LLMs across over 800 games, testing basic understanding, move selection, and puzzle solving. Results reveal significant shortcomings: no model beats Maia-1100 (human amateur level), and some lose to random play. We also present a strong baseline: our fine-tuned Qwen3-8B substantially improves performance, approaching much larger state-of-the-art reasoning models.

Reinforcement learning (RL) has become a powerful paradigm for optimizing large language models (LLMs) to handle complex reasoning tasks. A core challenge in this process lies in managing policy entropy, which reflects the balance between exploration and exploitation during training. Existing methods, such as proximal policy optimization (PPO) and its variants, discard valuable gradient signals from low-probability tokens due to the clipping mechanism. We systematically analyze the entropy dynamics and reveal that these clipped tokens play a critical yet overlooked role in regulating entropy evolution. We propose \textbf{C}oordinating \textbf{E}ntropy via \textbf{G}radient-\textbf{P}reserving \textbf{P}olicy \textbf{O}ptimization (CE-GPPO), a novel algorithm that reintroduces gradients from clipped tokens in native PPO in a gentle and bounded manner. By controlling the magnitude of gradients from tokens outside the clipping interval, CE-GPPO is able to achieve an exploration-exploitation trade-off. We provide theoretical justification and empirical evidence showing that CE-GPPO effectively mitigates entropy instability. Extensive experiments on mathematical reasoning benchmarks show that CE-GPPO consistently outperforms strong baselines across different model scales.

Foundation models excel across diverse tasks, but adapting them to specialized applications often requires fine-tuning, an approach that is memory and compute-intensive. Parameter-efficient fine-tuning (PEFT) methods mitigate this by updating only a small subset of weights. In this paper, we introduce HyperAdapt, a parameter-efficient fine-tuning method that significantly reduces the number of trainable parameters compared to state-of-the-art methods like LoRA. Specifically, HyperAdapt adapts a pre-trained weight matrix by applying row- and column-wise scaling through diagonal matrices, thereby inducing a high-rank update while requiring only $n+m$ trainable parameters for an $n \times m$ matrix. Theoretically, we establish an upper bound on the rank of HyperAdapt's updates, and empirically, we confirm that it consistently induces high-rank transformations across model layers. Experiments on GLUE, arithmetic reasoning, and commonsense reasoning benchmarks with models up to 14B parameters demonstrate that HyperAdapt matches or nearly matches the performance of full fine-tuning and state-of-the-art PEFT methods while using orders of magnitude fewer trainable parameters.

Balancing spectral, spatial, and temporal resolutions is a key challenge in spectral imaging. The Dual-Camera Coded Aperture Snapshot Spectral Imaging (DC-CASSI) system alleviates this trade-off but suffers from severely ill-posed reconstruction problems due to its high compression ratio. Existing methods are constrained by scene-specific tuning or excessive reliance on paired training data. To address these issues, we propose a Total Variation (TV) subgradient-guided multi-source fusion framework for DC-CASSI reconstruction, comprising three core components: (1) An end-to-end Single-Disperser CASSI (SD-CASSI) observation model based on the tensor-form Kronecker $δ$, which establishes a rigorous mathematical foundation for physical constraints while enabling efficient adjoint operator implementation; (2) An adaptive spatial reference generator that integrates SD-CASSI's physical model and RGB subspace constraint, generating the reference image as reliable spatial prior; (3) A TV subgradient-guided regularization term that encodes local structural directions from the reference image into spectral reconstruction, achieving high-quality fused results. The framework is validated on simulated datasets and real-world datasets. Experimental results demonstrate that it achieves state-of-the-art reconstruction performance and robust noise resilience. This work not only establishes an interpretable theoretical foundation for subgradient-guided fusion but also provides a practical fusion-based paradigm for high-fidelity spectral image reconstruction in DC-CASSI systems. Source code: https://github.com/bestwishes43/ADMM-TVDS.

Automating scientific poster generation requires hierarchical document understanding and coherent content-layout planning. Existing methods often rely on flat summarization or optimize content and layout separately. As a result, they often suffer from information loss, weak logical flow, and poor visual balance. We present PosterForest, a training-free framework for scientific poster generation. Our method introduces the Poster Tree, a structured intermediate representation that captures document hierarchy and visual-textual semantics across multiple levels. Building on this representation, content and layout agents perform hierarchical reasoning and recursive refinement, progressively optimizing the poster from global organization to local composition. This joint optimization improves semantic coherence, logical flow, and visual harmony. Experiments show that PosterForest outperforms prior methods in both automatic and human evaluations, without additional training or domain-specific supervision.

While transfer learning is an effective strategy, it often overlooks the opportunity to leverage knowledge from numerous available models online. Addressing this multi-source transfer learning problem is a promising path to boost adaptability and cut re-training costs. However, existing methods remain inherently coarse-grained: they lack the precision needed for fine-grained knowledge extraction as well as the scalability required to aggregate knowledge from either large numbers of source models or models with high parameter counts. We address these limitations by leveraging Singular Value Decomposition (SVD) to first decompose each source model into its elementary, rank-one components. A subsequent aggregation stage then selects only the most salient components from all sources, thereby overcoming the previous efficiency and precision limitations. To best preserve and leverage the synthesized knowledge base, our method adapts to the target task by fine-tuning only the principal singular values of the merged matrix. In essence, this process recalibrates the importance of top SVD components. The proposed framework allows for efficient and scalable multi-source transfer learning in both vision and language domains, while remaining robust to perturbations in both the input space and the parameter space.

Recent Large Language Model (LLM)-based AutoML systems demonstrate impressive capabilities but face significant limitations such as constrained exploration strategies and a severe execution bottleneck. Exploration is hindered by one-shot methods lacking diversity and Monte Carlo Tree Search (MCTS) approaches that fail to recombine strong partial solutions. The execution bottleneck arises from lengthy code validation cycles that stifle iterative refinement. To overcome these challenges, we introduce KompeteAI, a novel AutoML framework with dynamic solution space exploration. Unlike previous MCTS methods that treat ideas in isolation, KompeteAI introduces a merging stage that composes top candidates. We further expand the hypothesis space by integrating Retrieval-Augmented Generation (RAG), sourcing ideas from Kaggle notebooks and arXiv papers to incorporate real-world strategies. KompeteAI also addresses the execution bottleneck via a predictive scoring model and an accelerated debugging method, assessing solution potential using early stage metrics to avoid costly full-code execution. This approach accelerates pipeline evaluation 6.9 times. KompeteAI outperforms leading methods (e.g., RD-agent, AIDE, and Ml-Master) by an average of 3\% on the primary AutoML benchmark, MLE-Bench. Additionally, we propose Kompete-bench to address limitations in MLE-Bench, where KompeteAI also achieves state-of-the-art results

3D Scene Graphs integrate both metric and semantic information, yet their structure remains underutilized for improving path planning efficiency and interpretability. In this work, we present S-Path, a situationally-aware path planner that leverages the metric-semantic structure of indoor 3D Scene Graphs to significantly enhance planning efficiency. S-Path follows a two-stage process: it first performs a search over a semantic graph derived from the scene graph to yield a human-understandable high-level path. This also identifies relevant regions for planning, which later allows the decomposition of the problem into smaller, independent subproblems that can be solved in parallel. We also introduce a replanning mechanism that, in the event of an infeasible path, reuses information from previously solved subproblems to update semantic heuristics and prioritize reuse to further improve the efficiency of future planning attempts. Extensive experiments on both real-world and simulated environments show that S-Path achieves average reductions of 6x in planning time while maintaining comparable path optimality to classical sampling-based planners and surpassing them in complex scenarios, making it an efficient and interpretable path planner for environments represented by indoor 3D Scene Graphs. Code available at: https://github.com/snt-arg/spath_ros

We introduce EduCoder, a domain-specialized tool designed to support utterance-level annotation of educational dialogue. While general-purpose text annotation tools for NLP and qualitative research abound, few address the complexities of coding education dialogue transcripts -- with diverse teacher-student and peer interactions. Common challenges include defining codebooks for complex pedagogical features, supporting both open-ended and categorical coding, and contextualizing utterances with external features, such as the lesson's purpose and the pedagogical value of the instruction. EduCoder is designed to address these challenges by providing a platform for researchers and domain experts to collaboratively define complex codebooks based on observed data. It incorporates both categorical and open-ended annotation types along with contextual materials. Additionally, it offers a side-by-side comparison of multiple annotators' responses, allowing comparison and calibration of annotations with others to improve data reliability. The system is open-source, with a demo video available.

Large language models (LLMs) achieve impressive performance on complex mathematical benchmarks yet sometimes fail on basic math reasoning while generating unnecessarily verbose responses. In this paper, we present LLMThinkBench, a systematic benchmark and comprehensive empirical study to evaluate the efficiency of reasoning in LLMs, focusing on the fundamental tradeoff between accuracy and overthinking. First, we formalize the accuracy-verbosity tradeoff. Second, we introduce the Overthinking Score, a harmonic-mean metric combining accuracy and token-efficiency for holistic model evaluation. Third, we establish an evaluation protocol with dynamically-generated data across 14 basic math tasks. Fourth, we conduct a large-scale empirical study evaluating 53 LLMs, including reasoning and quantized variants across different reasoning budgets. Fifth, we release LLMThinkBench as an open-source Python package and public leaderboard for reproducibility. Our findings reveal: 1) model performance on complex benchmarks does not translate directly to basic math reasoning; 2) reasoning models generate ~18x more tokens while sometimes achieving lower accuracy and exhibit catastrophic collapse when tokens are constrained, dropping by up to ~36%; 3) the accuracy-verbosity relationship is non-monotonic with extended reasoning budgets yielding diminishing returns (GPT-5/o-series models show zero accuracy gain from low -> medium -> high reasoning effort). Our findings challenge the assumption that longer reasoning in LLMs necessarily improves mathematical reasoning. Our public leaderboard is available at https://ctrl-gaurav.github.io/LLMThinkBench/. Our open-source Python package is available at https://pypi.org/project/llmthinkbench/, and the codebase can be found at https://github.com/ctrl-gaurav/LLMThinkBench for easy and reproducible evaluation.

Multilingual Large Language Models considerably changed how technologies influence language. While previous technologies could mediate or assist humans, there is now a tendency to offload the task of writing itself to these technologies, enabling models to change our languages more directly. While they provide us quick access to information and impressively fluent output, beneath their (apparent) sophistication lies a subtle, insidious threat: the gradual decline and loss of linguistic diversity. In this position paper, I explore how model collapse, with a particular focus on translation technology, can lead to the loss of linguistic forms, grammatical features, and cultural nuance. Model collapse refers to the consequences of self-consuming training loops, where automatically generated data (re-)enters the training data, leading to a gradual distortion of the data distribution and the underrepresentation of low-probability linguistic phenomena. Drawing on recent work in Computer Vision, Natural Language Processing and Machine Translation, I argue that the many tails of our linguistic distributions might be vanishing, and with them, the narratives and identities they carry. This paper is a call to resist linguistic flattening and to reimagine Natural Language Processing as a field that encourages, values and protects expressive multilingual diversity and creativity.

Scaling test-time compute has emerged as an effective strategy for improving the performance of large language models. However, existing methods typically allocate compute uniformly across all queries, overlooking variation in query difficulty. To address this inefficiency, we formulate test-time compute allocation as a novel bandit learning problem and propose adaptive algorithms that estimate query difficulty on the fly and allocate compute accordingly. Compared to uniform allocation, our algorithms allocate more compute to challenging queries while maintaining accuracy on easier ones. Among challenging queries, our algorithms further learn to prioritize solvable instances, effectively reducing excessive computing on unsolvable queries. We theoretically prove that our algorithms achieve better compute efficiency than uniform allocation and empirically validate their effectiveness on math and code benchmarks. Specifically, our algorithms achieve up to an 11.10% performance improvement (15.04% relative) on the MATH-500 dataset, up to 10.82% (14.44% relative) on the AIME25 dataset, and up to an 11.23% performance improvement (15.29% relative) on the LiveCodeBench dataset.

The rapid development of Artificial Intelligence Generated Content (AIGC) has made high-fidelity generated audio widely available across the Internet, driving the advancement of audio steganography. Benefiting from advances in deep learning, current audio steganography schemes are mainly based on encoder-decoder network architectures. While these methods guarantee a certain level of perceptual quality for stego audio, they typically face high computational cost and long implementation time, as well as poor anti-steganalysis performance. To address the aforementioned issues, we pioneer a Fixed Decoder Network-Based Audio Steganography with Adversarial Perturbation Generation (FGAS). Adversarial perturbations carrying a secret message are embedded into the cover audio to generate stego audio. The receiver only needs to share the structure and key of the fixed decoder network to accurately extract the secret message from the stego audio. In FGAS, we propose an Audio Adversarial Perturbation Generation (A2PG) strategy with an optional robust extension and design a lightweight fixed decoder. The fixed decoder guarantees reliable extraction of the hidden message, while adversarial perturbations are optimized to keep the stego audio perceptually and statistically close to the cover audio, thereby improving anti-steganalysis performance. The experimental results show that FGAS significantly improves stego audio quality, achieving an average PSNR gain of over 10 dB compared to SOTA methods. Furthermore, FGAS demonstrates strong robustness against common audio processing attacks. Moreover, FGAS exhibits superior anti-steganalysis performance across different relative payloads; under high-capacity embedding, it achieves a classification error rate about 2% higher, indicating stronger anti-steganalysis performance than current SOTA methods.