Searcharxiv⌕ Search

SEARCH · Searcharxiv

Search Searcharxiv

Search indexed arXiv papers on artificial intelligence, large language models, computer vision and robotics. Read source abstracts and follow links to arXiv.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 145 records · Page 8Linked to original sources
arXiv2026

Factors Influencing the Emergence of Dependency Length Minimization in Neural Agent Simulations

Given various grammatical options, language users prefer the word order choice that reduces the overall length of syntactic dependencies, a principle known as dependency length minimization (DLM). The origins of this preference remain an open question, particularly whether it originates from constraints on efficient information processing. Computational simulations provide a powerful approach to identifying the factors influencing the emergence of linguistic phenomena. However, previous simulations of DLM have not examined realistic interaction contexts and have produced mixed results. The present study investigates the emergence of DLM in artificial languages using a recently proposed language learning and communication framework based on recurrent neural networks (RNNs). In this framework, agents are trained to speak and interpret artificial languages and then use these languages to communicate. Using this framework, we study the impact of several factors related to processing limitations in a communicative setting, such as noise during listening, limited speaker capacity, and incremental sentence processing. Our results reveal a complex interplay among these factors in shaping word order preferences in neural agents. Specifically, in the full meaning space, agents regularize toward a single dominant word order, while in the half meaning space they show a short-before-long preference that only aligns with DLM in verb-initial languages. A consistent DLM preference emerges only when agents are subject to incremental processing pressure. These findings suggest that limitations in human cognitive processing may indeed play a role in shaping DLM. Our findings provide insights into the conditions under which neural models replicate human-like preferences and highlight the challenges of designing emergent communication models that capture human cognitive biases in language processing.

cs.CL↗
arXiv2026

Abstention vs. Hallucination: Benchmarking LLM Source Attribution for Scientific Citations

Large language models (LLMs) increasingly generate citation-backed responses, yet citation hallucination remains a major challenge for trustworthy scientific information access. We introduce REASONS, a benchmark of 12,723 sentence-level citation instances spanning 12 arXiv subject categories, designed to evaluate scientific citation attribution under varying evidence conditions. We propose a dual-metric framework consisting of Abstention Rate (AR) and Hallucination Rate (HR) to characterize the trade-off between reliability and responsiveness. Using author-attribution and title-attribution tasks, we evaluate proprietary and open-source LLMs under zero-context, metadata-augmented, cascaded metadata-augmented prompting (CMP), retrieval-augmented, and adversarial settings. Advanced RAG lowers HR relative to Naive RAG (65.4% vs. 87.6%) but reduces AR from 5.0% to 0%. Under adversarial metadata, several systems exceed 85% HR, while retrieval-augmented variants frequently maintain near-zero abstention. Human evaluation of 1,000 outputs ($κ=0.78$) finds a 12.7:1 ratio of factual hallucinations to acceptable paraphrases. Our findings demonstrate that citation attribution systems should be evaluated not only for correctness but also for their ability to abstain appropriately under uncertainty. REASONS provides a benchmark and evaluation framework for studying attribution reliability in citation generation.

cs.CL↗
arXiv2026

Learning from Many Voices: Literary MT Using Multi-Reference Human and Synthetic Data

Unlike many other texts, literary works are often translated multiple times. We investigate strategies for leveraging these multi-reference datasets to improve literary machine translation. We propose a filtering framework based on semantic similarity to identify source texts whose references display meaningful variation while remaining faithful. We find that fine-tuning with medium to high semantic similarity data substantially outperforms low semantic similarity data. Moreover, using medium and high semantic similarity data achieves comparable or better performance than using the full unfiltered data. Synthetic translations generated by LLMs are economical and convenient alternatives to human expert translations; however, we find fine-tuning on human expert translations outperforms fine-tuning on synthetically augmented data in automatic metrics and human evaluations, demonstrating the indispensable value of human expert translations for fine-tuning literary machine translation models.

cs.CL↗
arXiv2026

Alignment Whack-a-Mole : Finetuning Activates Verbatim Recall of Copyrighted Books in Large Language Models

Frontier LLM companies have repeatedly assured courts and regulators that their models do not store copies of training data. They further rely on safety alignment strategies via RLHF, system prompts, and output filters to block verbatim regurgitation of copyrighted works, and have cited the efficacy of these measures in their legal defenses against copyright infringement claims. We show that finetuning bypasses these protections: by training models to expand plot summaries into full text, a task naturally suited for commercial writing assistants, we cause GPT-4o, Gemini-2.5-Pro, and DeepSeek-V3.1 to reproduce up to 85-90% of held-out copyrighted books, with single verbatim spans exceeding 460 words, using only semantic descriptions as prompts and no actual book text. This extraction generalizes across authors: finetuning exclusively on Haruki Murakami's novels unlocks verbatim recall of copyrighted books from over 30 unrelated authors. The effect is not specific to any training author or corpus: random author pairs and public-domain finetuning data produce comparable extraction, while finetuning on synthetic text yields near-zero extraction, indicating that finetuning on individual authors' works reactivates latent memorization from pretraining. Three models from different providers memorize the same books in the same regions ($r \ge 0.90$), pointing to an industry-wide vulnerability. Our findings offer compelling evidence that model weights store copies of copyrighted works and that the security failures that manifest after finetuning on individual authors' works undermine a key premise of recent fair use rulings, where courts have conditioned favorable outcomes on the adequacy of measures preventing reproduction of protected expression.

cs.CL↗
arXiv2026

IndicQE-APE: A Consolidated Benchmark for Quality Estimation and Automatic Post-Editing for Indic Languages

Indic quality estimation (QE) and automatic post-editing (APE) data is spread across separate releases, so no single resource supports training and evaluation across tasks and language pairs on one footing. We consolidate the WMT 2020-2024 shared-task lineage with an extended English-Malayalam resource into IndicQE-APE: $126{,}754$ instances over nine directional pairs, with up to four label types aligned on the same segment, a direct assessment, a human post-edit, word-level tags and an error explanation, and a test set stratified over four difficulty axes. We benchmark six prompted LLMs and three COMET metrics on segment-level QE, and three systems on APE. Two of the axes are defined partly on direct assessment and select a compressed slice of it. Segments whose segment-level and token-level signals disagree are ranked below equally scored segments of the same language. Four-shot prompting costs every model at or below $3.4$B both correlation and output-format compliance. Unedited MT beats every APE system we run on three of the four pairs. The benchmark (https://huggingface.co/datasets/surrey-nlp/IndicQE-APE) and code (https://github.com/surrey-nlp/IndicQE-APE) are released.

cs.CL↗
arXiv2026

Same Problem, Different Field: Cross-Domain Solution Import via Domain-Stripped Computational Fingerprints

The same underlying computational problem is solved across unrelated fields under different names: recursive Bayesian state estimation appears as a "Kalman filter" in control, "Bayesian forecasting" in pharmacokinetics, and "data assimilation" in geoscience. Topical and citation-based scientific embeddings cannot see this shared problem. We distill each paper once into a domain- and method-name-stripped faceted computational fingerprint, a free-text mechanism skeleton plus controlled computational facets. We define a tunable, facet-selectable similarity over it. The goal is solution import: surface cross-field pairs solving the same problem, so a bespoke implementation can be swapped for another field's standard, specialized solver. On a benchmark of 18 method families across 109 papers, the skeleton lifts cross-domain retrieval average precision over the abstract from 0.222 to 0.513, and the whole fingerprint reaches 0.557. Strikingly, four trained scientific embedders all fall below plain abstract+TF-IDF: they encode topical and citation similarity, the wrong signal for this task. The gain is the representation: the abstract-to-skeleton swap lifts every embedder, and the pipeline is one cached LLM call per paper plus a cheap embedder. An interventional re-skin / math-edit test shows the fingerprint tracks the computation, not the field. On a 501-paper wild corpus, known twins dominate the top of the ranking (23 of the top 30); with planted pairs excluded from the results, three blind LLM judges rate 3 of the top 5 and 8 of the top 30 pairs genuine import candidates, and 0 of 30 random ones. The human verification is the four executed imports: in one, an open standard solver reproduces a bespoke clinical dosing engine's output. We release the benchmark, the code, and the distillation prompt.

cs.DL↗
arXiv2026

EviSI: An Evidence-Based Evaluation Agent for Simultaneous Interpreting

Low-latency simultaneous speech-to-speech translation must keep pace with ongoing speech while preserving key information. To meet these demands, systems use segmentation, reformulation and condensation to reorganize and rephrase information. However, metrics developed for text translation, including BLEU and COMET, may not consistently distinguish faithful adaptations from semantic errors. We propose EviSI, a large language model evaluation agent combining Multidimensional Quality Metrics (MQM) with criteria developed with professional interpreters. Shared source evidence guides assessment across four dimensions: Anchor, Event, Logic and Fluency. Verified errors are deduplicated before deterministic scoring. On human-rated English to Chinese and Chinese to English data, EviSI recovers the aggregate English to Chinese human system ranking. Mean within-dataset Kendall correlations for system rankings reach 0.707 and 0.467, respectively, exceeding evaluated BLEU and COMET baselines. A multilingual extension to five directions without human ratings retains the dimensions and scoring rule, showing positive system ranking correlations with COMET throughout.

cs.CL↗
arXiv2026

EdiTikZ: Scientific Figure Editing from Revision Trajectories

Vision-language models (VLMs) have shown strong performance in generating scientific figures from text or images. However, publication-ready figures often require iterative refinement, making scientific figure editing an important yet largely unexplored step toward interactive figure creation. Existing approaches rely on costly proprietary agentic systems, focus primarily on evaluation, or construct training supervision from synthetically generated edits. Instead, we leverage naturally occurring scientific revision and development trajectories as a scalable source of supervision. To this end, we introduce DaEdiTikZ, the first large-scale dataset of revision-derived scientific figure edits, constructed by mining 391K plausible TikZ edit pairs from arXiv, GitHub, and TeX SE and inferring 781K directed edit instructions with a VLM conditioned on rendered figures and TikZ code. We further introduce DaEdiTikZ-Bench, a human-refined benchmark with 690 instances, and train two compact Qwen3.5-based EdiTikZ models (4B and 9B) by jointly learning image-to-TikZ reconstruction and instruction-conditioned editing, followed by reinforcement learning (RL) with complementary rewards for rendered fidelity and edit application. Automatic evaluation places our 9B model above all tested baselines, while human evaluation with 9 annotators and 4,320 ratings places it above GPT-5.6-Sol and on par with Gemini-3.1-Pro. Under severe out-of-distribution shifts, it remains competitive with GPT-5.6-Sol near its 2K training sequence-length regime.

cs.AI↗
arXiv2026

A rigor-matched audit of periodic-step layer skipping for efficient llm inference: conflayers versus swift, with a supplemental analysis of trained routing alternatives

Layer-skipping methods for efficient LLM inference decide, at some granularity, which transformer layers to execute for a given input. We present a rigor-matched, three-seed audit of two periodic-step, search-based methods that make this decision online at inference time and re-evaluate it every few generation steps: a confidence-gated early-exit baseline (ConfLayers) and genuine self-speculative decoding (SWIFT, Xia et al. 2024), together with vanilla autoregressive decoding, across Qwen2.5-0.5B and Qwen2.5-1.5B (Yang et al. 2024) on GSM8K (Cobbe et al. 2021) and CNN/DailyMail (Nallapati et al. 2016; See et al. 2017). SWIFT is strongest on accuracy in three of four cells, while ConfLayers is dominated everywhere, with particularly large deficits on GSM8K at 1.5B. Separating online-search overhead from pure inference cost, we find that SWIFT's true inference speed is faster than ConfLayers's in all four cells (5-21%), reversing the naive wall-clock ranking in three; ConfLayers's search overhead is small and stable (1-2% of cost), whereas SWIFT's is larger and more variable across seeds (up to 28.7%). We additionally examine two trained routing methods, LayerRoute (Sikdar, 2026), a per-sequence, input-conditioned hard gate, and LayerDrop (Fan et al. 2020), a fixed, input-independent pruning pattern, as a supplemental analysis rather than a head-to-head comparison because both operate at a coarser decision granularity. Under a verified protocol with genuine per-input gating, a genuine full-model baseline, and genuine inference-time compute skipping, both show modest real speedups (1.08-1.33x) but accuracy well below the periodic-step methods, including near-total LayerRoute collapse on GSM8K at 1.5B (0.003 mean exact-match across three seeds). We release the full audit protocol as a template for rigor-matched efficiency comparisons.

cs.CL↗
arXiv2026

VoiceCodeBench: Evaluating Exact Structured-Token Recovery in Automatic Speech Recognition

Automatic speech recognition is usually evaluated with word error rate (WER), although voice workflows often require exact written values. VoiceCodeBench measures whether transcripts preserve identifiers, paths, commands, and other structured tokens needed by downstream software. It contains 300 human-recorded English workplace segments (5.59 hours, 85 speakers) and 1,482 audited entities across 26 types and eight domains. Under a raw-audio-only protocol, we evaluate 19 batch and streaming systems using WER, Canonical Token/Entity Match (CTEM), and strict segment-level Task Success Rate (TSR). Across systems, WER has little rank agreement with CTEM (Spearman $ρ=-0.28$) or TSR ($ρ=-0.22$). The best CTEM and TSR are 91.8% and 68.7%. Even the strongest system therefore leaves nearly one-third of recordings with an unrecovered critical value. Symbol-, separator-, and boundary-sensitive entities account for most errors.

cs.CL↗
arXiv2026

Harnessing the Reasoning Economy: A Survey of Efficient Reasoning for Large Language Models

Recent advancements in Large Language Models (LLMs) have significantly enhanced their ability to perform complex reasoning tasks, transitioning from fast and intuitive thinking (System 1) to slow and deep reasoning (System 2). While System 2 reasoning improves task accuracy, it often incurs substantial computational costs due to its slow thinking nature and inefficient or unnecessary reasoning behaviors. In contrast, System 1 reasoning is computationally efficient but leads to suboptimal performance. Consequently, it is critical to balance the trade-off between performance (benefits) and computational costs (budgets), giving rise to the concept of reasoning economy. In this survey, we provide a comprehensive analysis of reasoning economy in both the post-training and test-time inference stages of LLMs, encompassing i) the cause of reasoning inefficiency, ii) behavior analysis of different reasoning patterns, and iii) potential solutions to achieve reasoning economy. By offering actionable insights and highlighting open challenges, we aim to shed light on strategies for improving the reasoning economy of LLMs, thereby serving as a valuable resource for advancing research in this evolving area. We also provide a public repository to continually track developments in this fast-evolving field.

cs.CL↗
arXiv2026

Evaluating the Scalability and Adversarial Generalization of GRPO-Trained NLI Models

Natural Language Inference (NLI) is a central task in natural language understanding with applications in fact-checking, question answering, and information retrieval. Despite its importance, current NLI systems heavily rely on supervised learning with datasets that often contain annotation artifacts and biases, limiting generalization and real-world applicability. In this work, we apply a reinforcement learning-based approach using Group Relative Policy Optimization (GRPO) for Chain-of-Thought (CoT) learning in NLI, eliminating the need for human-labeled rationales and enabling this type of training on challenging datasets such as ANLI. We fine-tune 7B, 14B, and 32B language models using parameter-efficient techniques (LoRA and QLoRA), demonstrating strong performance across standard and adversarial NLI benchmarks. At the 32B scale, GRPO-trained models generalize better than other supervised baselines in adversarial sets. With AWQ quantization, the 32B model fits within 22GB of CUDA memory. This work provides a scalable and practical framework for building robust NLI systems without sacrificing inference quality.

cs.CL↗
arXiv2026

ID-Align: RoPE-Conscious Position Remapping for Dynamic High-Resolution Adaptation in Vision-Language Models

Currently, a prevalent approach for enhancing Vision-Language Models (VLMs) performance is to encode both the high-resolution version and the thumbnail of an image simultaneously. While effective, this method generates a large number of image tokens. When combined with the widely used Rotary Position Embedding (RoPE), its long-term decay property hinders the interaction between high-resolution tokens and thumbnail tokens, as well as between text and image. To address these issues, we propose ID-Align, which alleviates these problems by reordering position IDs. In this method, high-resolution tokens inherit IDs from their corresponding thumbnail token while constraining the overexpansion of positional indices. Our experiments conducted within the LLaVA-Next framework demonstrate that ID-Align achieves significant improvements, including a 6.09% enhancement on MMBench's relation reasoning tasks and notable gains across multiple benchmarks. Our code is available at the following link: https://github.com/zooblastlbz/ID-Align.

cs.CV↗
arXiv2026

KARE-RAG: Knowledge-Aware Refinement and Enhancement for RAG

Retrieval-Augmented Generation (RAG) equips large language models with external knowledge and is central to knowledge-intensive tasks. As RAG systems enter real-world use, generators must reliably leverage retrieved evidence. Recent fine-tuning methods improve adaptation to RAG scenarios, but optimization remains challenging because retrieval may return incomplete, fragmented, noisy, or conflicting contexts. Complex tasks further require fine-grained evidence dependencies. These challenges make high-quality supervision costly and limit generalization. We present KARE-RAG (Knowledge-Aware Refinement and Enhancement for RAG), a training-time scaffolded alignment framework. It uses structured knowledge representations as temporary scaffolds to expose evidence organization, support localized factual refinement, and construct fine-grained preference pairs. In our main implementation, an expert LLM refines a lightweight graph-structured evidence sketch. The generator is optimized with token-weighted Dense Direct Preference Optimization (DDPO), which focuses learning on edited scaffold regions. Scaffolds are used only for data construction and training supervision. At inference time, the model runs as standard Vanilla RAG without graph construction, extra retrieval, or latency overhead. Experiments show that KARE improves transfer across the evaluated QA and relation extraction datasets with limited training data, while leaving general capabilities largely unchanged. KARE can also complement existing RAG training objectives as an additional alignment stage.

cs.CL↗
arXiv2026

CogniDir: Combating Cognitive Malicious Comments via Adaptive Distributional Learning for Robust Fake News Detection

The proliferation of Large Language Models (LLMs) has enabled a new class of psychologically grounded malicious comments, shifting fake news attacks from surface-level textual noise to deep cognitive and logical manipulation. This shift severely undermines existing detectors, which conventionally rely on static attack assumptions and fixed training distributions. To bridge this gap, we introduce CogniDir, an adaptive distributional learning framework that reformulates robust detection as a dynamic data mixture optimization problem for social media content safety. Grounded in cognitive psychology, we first formalize mechanism-specific cognitive adversarial paradigms to systematically expose deep-seated detector vulnerabilities. To address the vulnerability heterogeneity, CogniDir derives an information-theoretic score coupling empirical accuracy with probabilistic confidence, which is then mapped to adaptive sampling proportions through a Dirichlet-mean parameterization. This formulation enables smooth, feedback-driven reallocation of training exposure toward the most brittle attack mechanisms. Experimental results on three benchmarks demonstrate that CogniDir yields state-of-the-art robustness, improving F1 scores by up to 17.9% over competitive baselines under heterogeneous, AI-generated adversarial pressures.

cs.LG↗
arXiv2026

From Representation to Enactment: The ABC Framework of the Translating Mind

Building on the third-wave Extended Mind (EM) theory and radical enactivism, this article suggests an alternative to representation-based models of the mind. We build on the ABC framework in which translation is not understood as the manipulation of static interlingual correspondences but an enacted activity, dynamically integrating affective-evaluative, behavioral-enacting, and cognitive-inferential (ABC) processes. Drawing on affordance theory, we argue that the translating mind is the dynamically organized process through which the translator and its environment constitute and transform a landscape of enlanguaged translation affordances for action, that emerges through loops of brain-body-environment interactions. This non-representational account reframes translation as skillful participation in sociocultural practice, where meaning is co-created in real time through embodied interaction with texts, tools, and contexts.

cs.CL↗
arXiv2026

CoT is Not the Chain of Truth: An Empirical Internal Analysis of Reasoning LLMs for Fake News Generation

From generating headlines to fabricating news, the Large Language Models (LLMs) are typically assessed by their final outputs, under the safety assumption that a refusal response signifies safe reasoning throughout the entire process. Challenging this assumption, our study reveals that during fake news generation, even when a model rejects a harmful request, its Chain-of-Thought (CoT) reasoning may still internally contain and propagate unsafe narratives. To analyze this phenomenon, we introduce a unified safety-analysis framework that systematically deconstructs CoT generation across model layers and evaluates the role of individual attention heads through Jacobian-based spectral metrics. Within this framework, we introduce three interpretable measures: stability, geometry, and energy to quantify how specific attention heads respond or embed deceptive reasoning patterns. Extensive experiments on multiple reasoning-oriented LLMs show that the generation risk rises significantly when the thinking mode is activated, where the critical routing decisions are concentrated in only a few contiguous mid-depth layers. By precisely identifying the attention heads responsible for this divergence, our work challenges the assumption that refusal implies safety and provides a new understanding perspective for mitigating latent reasoning risks.

cs.CL↗
arXiv2026

Explainable Token-level Noise Filtering for LLM Fine-tuning Datasets

Large Language Models (LLMs) have seen remarkable advancements, achieving state-of-the-art results in diverse applications. Fine-tuning, an important step for adapting LLMs to specific downstream tasks, typically involves further training on corresponding datasets. However, a fundamental discrepancy exists between current fine-tuning datasets and the token-level optimization mechanism of LLMs: most datasets are designed at the sentence-level, which introduces token-level noise, causing negative influence to final performance. In this paper, we propose XTF, an explainable token-level noise filtering framework. XTF decomposes the complex and subtle contributions of token-level data to the fine-tuning process into three distinct and explicit attributes (reasoning importance, knowledge novelty, and task relevance), which can be assessed using scoring methods, and then masks the gradients of selected noisy tokens accordingly to optimize the performance of fine-tuned LLMs. We conduct extensive experiments on three representative downstream tasks (math, code and medicine) across 7 mainstream LLMs. The results demonstrate that XTF can significantly improve downstream performance by up to 13.7% compared to regular fine-tuning. Our work highlights the importance of token-level dataset optimization, and demonstrates the potential of strategies based on attribute decomposition for explaining complex training mechanisms.

cs.CL↗