AI conference papers

AI conference paper library

Search papers from 30+ top conferences, including NeurIPS, ICLR, ICML, CVPR, ACL and AAAI. Filter by keyword, then open any paper for its abstract, authors and links.

arXiv

CARE: Anti-entanglement Ultrasound Image Segmentation via Channel-Aware Region Extrication

Weixin Xu, Yuting Lu, Luqi Gong, Qing Guo, Ziliang Wang, Yun Xing

Accurate ultrasound image segmentation is fundamentally challenged by target-context entanglement, where lesion cues are easily mixed with surrounding tissues and artifacts of similar appearance. Although existing methods often localize suspicious regions reasonably well, they remain vulnerable to ambiguous predictions because they mainly strengthen feature extraction or context aggregation, rather than explicitly organizing how lesion and interference cues are represented and distinguished. To address this limitation, we propose Channel-Aware Region Extrication (CARE), a segmentation framework that improves ultrasound segmentation by progressively extricating lesion evidence from visually entangled context. Instead of merely reweighting features, CARE explicitly separates encoded responses according to their lesion relevance and then re-evaluates the resulting complementary representations through reciprocal region interaction, so that suppressed lesion cues can be recovered while misleading contextual activations are corrected. In this way, CARE promotes target-context discrimination directly in the learned representation, without sacrificing localization quality. Extensive experiments on BUSI, BUSIS, and TN3K benchmarks show that CARE consistently achieves superior performance, thereby validating representation extrication as an effective solution for addressing the inherent visual ambiguity in ultrasound segmentation.

arXiv

Latent Interpolation Learning Using Diffusion Models for Cardiac Volume Reconstruction

Niklas Bubeck, Suprosanna Shit, Chen Chen, Can Zhao, Pengfei Guo, Dong Yang, Georg Zitzlsberger, Daguang Xu, Bernhard Kainz, Daniel Rueckert, Jiazhen Pan

Cardiac Magnetic Resonance (CMR) imaging is a critical tool for diagnosing and managing cardiovascular disease, yet its utility is often limited by the sparse acquisition of 2D short-axis slices, resulting in incomplete volumetric information. Accurate 3D reconstruction from these sparse slices is essential for comprehensive cardiac assessment, but existing methods face challenges, including reliance on predefined interpolation schemes (e.g., linear or spherical), computational inefficiency, and dependence on additional semantic inputs such as segmentation labels or motion data. To address these limitations, we propose a novel Cardiac Latent Interpolation Diffusion (CaLID) framework that introduces three key innovations. First, we present a data-driven interpolation scheme based on diffusion models, which can capture complex, non-linear relationships between sparse slices and improves reconstruction accuracy. Second, we design a computationally efficient method that operates in the latent space and speeds up 3D whole-heart upsampling time by a factor of 24, reducing computational overhead compared to previous methods. Third, with only sparse 2D CMR images as input, our method achieves SOTA performance against baseline methods, eliminating the need for auxiliary input such as morphological guidance, thus simplifying workflows. We further extend our method to 2D+T data, enabling the effective modeling of spatiotemporal dynamics and ensuring temporal coherence. Extensive volumetric evaluations and downstream segmentation tasks demonstrate that CaLID achieves superior reconstruction quality and efficiency. By addressing the fundamental limitations of existing approaches, our framework advances the state of the art for spatio and spatiotemporal whole-heart reconstruction, offering a robust and clinically practical solution for cardiovascular imaging.

arXiv

Joint Lossless Compression and Steganography for Medical Images via Large Language Models

Pengcheng Zheng, Xiaorong Pu, Kecheng Chen, Jiaxin Huang, Meng Yang, Bai Feng, Yazhou Ren, Jianan Jiang, Chaoning Zhang, Yang Yang, Heng Tao Shen

Recently, large language models (LLMs) have driven promising progress in lossless image compression. However, directly adopting existing paradigms for medical images suffers from an unsatisfactory trade-off between compression performance and efficiency. Moreover, existing LLM-based compressors often overlook the security of the compression process, which is critical in modern medical scenarios. To this end, we propose a novel joint lossless compression and steganography framework. Inspired by bit plane slicing (BPS), we find it feasible to securely embed privacy messages into medical images in an invisible manner. Based on this insight, an adaptive modalities decomposition strategy is first devised to partition the entire image into two segments, providing global and local modalities for subsequent dual-path lossless compression. During this dual-path stage, we innovatively propose a segmented message steganography algorithm within the local modality path to ensure the security of the compression process. Coupled with the proposed anatomical priors-based low-rank adaptation (A-LoRA) fine-tuning strategy, extensive experimental results demonstrate the superiority of our proposed method in terms of compression ratios, efficiency, and security. The source code will be made publicly available.

arXiv

Improving Large Vision-Language Models' Understanding for Flow Field Data

Xiaomei Zhang, Hanyu Zheng, Xiangyu Zhu, Jinghuan Wei, Junhong Zou, Zhen Lei, Zhaoxiang Zhang

Large Vision-Language Models (LVLMs) have shown impressive capabilities across a range of tasks that integrate visual and textual understanding, such as image captioning and visual question answering. These models are trained on large-scale image and video datasets paired with text, enabling them to bridge visual perception and natural language processing. However, their application to scientific domains, especially in interpreting complex field data commonly used in the natural sciences, remains underexplored. In this work, we introduce FieldLVLM, a novel framework designed to improve large vision-language models' understanding of field data. FieldLVLM consists of two main components: a field-aware language generation strategy and a data-compressed multimodal model tuning. The field-aware language generation strategy leverages a special-purpose machine learning pipeline to extract key physical features from field data, such as flow classification, Reynolds number, and vortex patterns. This information is then converted into structured textual descriptions that serve as a dataset. The data-compressed multimodal model tuning focuses on LVLMs with these generated datasets, using a data compression strategy to reduce the complexity of field inputs and retain only the most informative values. This ensures compatibility with the models language decoder and guides its learning more effectively. Experimental results on newly proposed benchmark datasets demonstrate that FieldLVLM significantly outperforms existing methods in tasks involving scientific field data. Our findings suggest that this approach opens up new possibilities for applying large vision-language models to scientific research, helping bridge the gap between large models and domain-specific discovery.

arXiv

NWaaS: A Non-Intrusive and Privacy-Preserving Watermarking-as-a-Service System with Adaptive Resource Scheduling

Haonan An, Qianyao Ren, Guang Hua, Tao Li, Yu Guo, Yanan Ma, Hangcheng Cao, Yuguang Fang

Securing intellectual property (IP) in Machine Learning as a Service is critical yet challenging. While deep neural network watermarking serves as a standard defense against model extraction, existing Watermarking-as-a-Service paradigms face a triple challenge of intrusiveness, privacy risks, and inefficiency. To address these challenges, we propose Non-intrusive Watermarking as a Service (NWaaS), a holistic framework enabling trustworthy and efficient IP protection. We first introduce $\mathtt{ShadowMark}$, a novel watermarking algorithm that establishes a side-channel for ownership verification without modifying the model. It ensures zero performance degradation and eliminates the need for parameter-heavy fine-tuning as well as access to original training data, thereby addressing the intrusiveness and inefficiency inherent in existing approaches. Leveraging this non-intrusive property, we design a collaborative partitioning mechanism that allows model owners to offload self-defined partial layers, enabling a flexible trade-off between IP privacy and service cost. Furthermore, to mitigate latency from collaborative computing under high concurrency and enhance system resource utilization, we propose proportion disparity joint scheduling, a payload-balancing resource scheduling algorithm tailored to the heterogeneous constraints of edge-cloud environments. Extensive experiments demonstrate that NWaaS provides robust ownership verification across diverse continuous X-to-Image modalities, while ensuring secure owner privacy protection and superior system performance.

arXiv

Toward Automated Detection of Documentation Inconsistencies in Electronic Health Records

Jian Lu, Panyu Chen, Miriam Treggiari, Robert Blessing, Danyang Zhuo, Chunhua Weng, William W. Stead, Anru R. Zhang

Objective: To characterize the kinds of internal documentation inconsistencies a general-domain large language model (LLM) can surface from real-world discharge summaries, and to identify recurring failure modes that limit reliability at scale. Materials and Methods: We applied a two-stage LLM pipeline---open-ended candidate identification (Gemini 2.5 Pro) followed by context-grounded verification (Gemini 2.5 Flash)---to 3,000 randomly sampled MIMIC-IV-Note discharge summaries. A subset of the pipeline output was then reviewed manually by clinical experts. Results: Our pipeline surfaced 3,460 candidate inconsistencies, affecting 69.7% of admissions. Representative examples spanned demographics, allergies, procedures, diagnoses, laboratory, medications, and care-planning domains, with direct implications for clinical reasoning or patient safety. Expert review also revealed recurring failure modes that arise when verification requires temporal reasoning, evolving-diagnosis context, or knowledge of outpatient-prescribing conventions the model does not natively possess. Discussion: Detection is highly context-dependent: many flagged pairs require anchoring each statement to its source section and clinical domain, then assessing whether the conflict reflects a true contradiction or missing context. We propose a graded ontology spanning strict contradiction and ambiguity, with a schema characterizing each flagged case by category, section, domain, and inconsistency axis. Conclusion: This formative study establishes a methodological foundation and conceptual framework to guide subsequent validated, large-scale EHR-inconsistency analysis.

arXiv

Not All LLM Reasoning is Visible in the Chain-of-Thought

Vatsal Baherwani, Tom Goldstein, Ashwinee Panda

A key question for AI safety is whether a language model expresses all of its reasoning in its output tokens. We demonstrate a concrete failure mode where frontier models exhibit invisible reasoning by leveraging semantically irrelevant filler tokens to improve performance on synthetic reasoning tasks. We evaluate 13 frontier language models across three tasks and find that many models benefit significantly from filler tokens, with accuracy improvements of up to 13 percentage points. The benefit depends on which tokens are used and differs across models. We further show that filler tokens enable Claude Opus 4.5 to satisfy a hidden modular arithmetic constraint without sacrificing accuracy on its primary task, demonstrating that invisible reasoning can serve objectives entirely invisible to CoT monitoring. Reinforcement learning gives Qwen3-235B strong preferences over filler token content, but neither RL nor supervised fine-tuning produces a filler token benefit that persists at test time. Our results indicate that frontier models already perform consequential computation with no interpretable trace in their output tokens.

arXiv

Simple Language Normalization Wins: Cross-Lingual Speaker Verification for the TidyVoice 2026 Challenge

Nina Hosseini-Kivanani

Cross-lingual mismatch remains a key source of overall degradation in modern speaker verification. The TidyVoice2026 Challenge targets this setting with text-independent verification, comprising 3,666 training and 808 development speakers in 40 languages and 2,200 evaluation speakers in 38 unseen languages, without language labels at test time. Starting from the official SimAM-ResNet34 baseline pretrained on VoxBlink2 and VoxCeleb2 and fine-tuned on TidyVoice, we revisit Nuisance Attribute Projection (NAP) as a simple language-normalization step in the embedding space. We estimate a compact language subspace from cross-language same-speaker differences and project embeddings onto its orthogonal complement before cosine scoring with Adaptive Symmetric score normalization. This reduces development EER from 2.97\% with cosine and 2.70\% with AS-Norm to 2.18\% and yields a Codabench evaluation score of 8.40, showing that simple back-end language normalization can rival more complex systems.

arXiv

AssumptionMiner: Extracting, Tracing, and Revising Implicit Assumptions in LLM Code Generation

Jie "JW" Wu

Large language models (LLMs) generate code from natural-language prompts, yet real-world prompts rarely provide complete specifications. When prompts leave input formats, error handling, or design decisions unspecified, LLMs fill these gaps with implicit assumptions that shape the generated code's behavior and correctness. Because these assumptions remain hidden, generated code may satisfy tests while violating developer intent. We present AssumptionMiner, a framework that makes implicit assumptions a first-class artifact of LLM-based code generation. In addition to code, AssumptionMiner produces an explicit assumption layer, a structured representation of inferred constraints and design decisions that developers can inspect, confirm, or revise. An AST-based dependency graph enables targeted regeneration of only the code affected by a revised assumption. We also introduce a benchmark of 180 ambiguous programming tasks with 676 annotated assumptions, including a human-verified subset for evaluating code localization. We evaluate assumption extraction, code localization, and assumption-guided regeneration. Across open-source LLMs, a confidence-weighted ensemble achieves an F1 score of 0.816 for assumption extraction, improving on the strongest offline baseline by 3.6x. On the human-verified localization benchmark, AST-guided localization identifies more precise code regions than keyword-based and whole-file baselines. During assumption revision, targeted regeneration modifies less code than non-targeted alternatives while exposing challenges in handling cascading edits. These results demonstrate that making assumptions explicit improves the transparency and controllability of LLM-based code generation.

arXiv

CHiPS: Character Histograms and Positional Signals for Lightweight Authorship Attribution in Romanian Texts

Sanda-Maria Avram, George C. Ţurcaş

We propose CHiPS, a lightweight character-level authorship attribution method for Romanian texts. All reported experiments are closed-set: the true author is one of the candidate authors in the training data. CHiPS studies two complementary fingerprints of writing style: CH-SVM, a character-histogram classifier based on one-character marginal distributions, and FFT12-LR, a positional-signal classifier that represents selected characters and punctuation classes as impulse trains (binary indicator sequences over character positions) and extracts Fourier/Welch spectral descriptors. We also report CHiPS-F, a leakage-safe decision-level fusion variant, and an optional top-5 listwise reranker trained only on out-of-fold predictions. The method requires no tokenization, syntactic analysis, pretrained language model, or transformer fine-tuning, and it avoids character $n$-gram features with $n \geq 2$ in the histogram component. On a locked grouped ROST split comprising 400 files from 392 source-text groups, written by 10 authors, with source-text-level evaluation and grouped five-fold model selection, CHiPS-F reaches 0.9310 accuracy and 0.9341 macro-F1. A matched but unrestricted character 2--5-gram TF--IDF SVM comparator reaches 1.0000 accuracy and macro-F1 on the same held-out groups, so the contribution is not a claim of best possible classification accuracy. Instead, the experiments ask how far restricted, transparent character evidence can go under strict leakage control. On ROSTories-cleaned, a secondary ROST-overlapping corpus comprising 1,248 files from 1,240 source-text groups, written by 19 authors, the same protocol gives 0.8919 accuracy and 0.8708 macro-F1 for CHiPS-R.

arXiv

Quality Text, Robust Vision: The Role of Language in Enhancing Visual Robustness of Vision-Language Models

Futa Waseda, Saku Sugawara, Isao Echizen

Defending pre-trained vision-language models (VLMs), such as CLIP, against adversarial attacks is crucial, as these models are widely used in diverse zero-shot tasks, including image classification. However, existing adversarial training (AT) methods for robust fine-tuning largely overlook the role of language in enhancing visual robustness. Specifically, (1) supervised AT methods rely on short texts (e.g., class labels) to generate adversarial perturbations, leading to overfitting to object classes in the training data, and (2) unsupervised AT avoids this overfitting but remains suboptimal against practical text-guided adversarial attacks due to its lack of semantic guidance. To address these limitations, we propose Quality Text-guided Adversarial Fine-Tuning (QT-AFT), which leverages high-quality captions during training to guide adversarial examples away from diverse semantics present in images. This enables the visual encoder to robustly recognize a broader range of image features even under adversarial noise, thereby enhancing robustness across diverse downstream tasks. QT-AFT overcomes the key weaknesses of prior methods -- overfitting in supervised AT and lack of semantic awareness in unsupervised AT -- achieving state-of-the-art zero-shot adversarial robustness and clean accuracy, evaluated across 16 zero-shot datasets. Furthermore, our comprehensive study uncovers several key insights into the role of language in enhancing vision robustness; for example, describing object properties in addition to object names further enhances zero-shot robustness. Our findings point to an urgent direction for future work -- centering high-quality linguistic supervision in robust visual representation learning.

arXiv

TDiR: Transformer based Diffusion for Image Restoration Tasks

Abbas Anwar, Ibrahim Radwan, Ali Arshad Nasir, Mudassir Masood, Saeed Anwar

Images captured in challenging environments often experience various types of degradation, such as noise, color cast, blur, and light scattering. These issues significantly lower image quality, thereby reducing their usefulness in downstream tasks such as object detection, mapping, and classification. Our transformer-based diffusion model was developed to address image restoration challenges and enhance the quality of degraded images. Our methodology is assessed across three primary image restoration tasks, including underwater enhancement, denoising, and deraining, utilizing five standard benchmarks. It is then compared to 18 state-of-the-art techniques, employing four evaluation metrics. Our results show that the diffusion model, combined with transformers, outperforms current methods. The findings highlight the effectiveness of diffusion models and transformers in improving degraded image quality, thereby broadening their application in downstream tasks that demand high-fidelity visual data

arXiv

Geometry-Guided Representations for Coherent Lane and Traffic Topology Reasoning in Driving Scenes

Yueru Luo, Changqing Zhou, Yiming Yang, Erlong Li, Chao Zheng, Shuguang Cui, Zhen Li

Road topology reasoning is fundamental for autonomous driving, requiring both accurate perception of road elements and understanding of their complex connectivity, including lane connectivity (Lane-to-Lane, L2L) and traffic regulation (Lane-to-Traffic signs, L2T). However, existing methods typically treat perception and topology reasoning as fragmented tasks, ignoring their potential for mutual enhancement. Crucially, while topology is inherently relational, prior works often overlook geometric relationships during feature extraction, relying instead on brittle post-processing or coordinate-based heuristics applied only at inference time. To bridge this gap, we propose CoPo (Coherent Perception and toPology), a unified framework that integrates geometry-guided relational modeling across three levels: 1) Perception-level: We introduce a relation-aware lane detector that utilizes geometry-biased self-attention and curve-guided cross-attention to enrich lane representations with structural priors; 2) Reasoning-level: We design relation-enhanced topology heads, including a geometry-enhanced L2L head and a cross-view L2T head, which effectively align features to infer connectivity; and 3) Supervision-level: We implement a contrastive InfoNCE strategy to regularize relational embeddings, pulling connected pairs closer in the latent space. This coherent multi-level design enables end-to-end joint optimization of perception and reasoning. Extensive experiments on OpenLane-V2 demonstrate that CoPo significantly outperforms existing methods, achieving gains of {\textbf{+3.1}} in DET$_l$, {\textbf{+5.3}} in TOP$_{ll}$, {\textbf{+4.9}} in TOP$_{lt}$, and {\textbf{+4.4}} overall in OLS, setting a new state-of-the-art.

arXiv

HunyuanVideo-HOMA: Generic Human-Object Interaction in Multimodal Driven Human Animation

Ziyao Huang, Zixiang Zhou, Juan Cao, Yifeng Ma, Yi Chen, Zejing Rao, Zhiyong Xu, Hongmei Wang, Qin Lin, Yuan Zhou, Qinglin Lu, Fan Tang

To address key limitations in human-object interaction (HOI) video generation -- specifically the reliance on curated motion data, limited generalization to novel objects/scenarios, and restricted accessibility -- we introduce HunyuanVideo-HOMA, a weakly conditioned multimodal-driven framework. HunyuanVideo-HOMA enhances controllability and reduces dependency on precise inputs through sparse, decoupled motion guidance. It encodes appearance and motion signals into the dual input space of a multimodal diffusion transformer (MMDiT), fusing them within a shared context space to synthesize temporally consistent and physically plausible interactions. To optimize training, we integrate a parameter-space HOI adapter initialized from pretrained MMDiT weights, preserving prior knowledge while enabling efficient adaptation, and a facial cross-attention adapter for anatomically accurate audio-driven lip synchronization. Extensive experiments confirm state-of-the-art performance in interaction naturalness and generalization under weak supervision. Finally, HunyuanVideo-HOMA demonstrates versatility in text-conditioned generation and interactive object manipulation, supported by a user-friendly demo interface. The project page is at https://bone-11.github.io/homa-page/.

arXiv

Hiding Faces in Plain Sight: Defending DeepFakes by Disrupting Face Detection

Delong Zhu, Yuezun Li, Baoyuan Wu, Jiaran Zhou, Zhibo Wang, Siwei Lyu

Face-swapping DeepFakes have become an escalating societal concern, attracting increasing attention in recent years. To counter this, we investigate a new proactive defense framework to prevent individuals from being victimized in DeepFake videos. The core idea of this framework is to contaminate the inputs of DeepFake models by disrupting face detectors, based on the observation that face detectors are commonly used to automatically extract victim faces in most DeepFake techniques. Once the face detectors malfunction, the faces will not be correctly extracted, thereby impairing the training or synthesis stages of DeepFake models. To achieve this, we describe a strategy named {\em FacePoison}, which fools face detectors by adding dedicated adversarial perturbations to video frames. Building upon this, we introduce {\em VideoFacePoison}, an extended strategy that can efficiently propagate FacePoison across video frames instead of applying it individually to each frame, thus significantly reducing the computational overhead while retaining favorable attack performance. This framework is validated on five face detectors, and extensive experiments against eleven different DeepFake models demonstrate the effectiveness of disrupting face detectors to hinder DeepFake generation. The source code is publicly available at: https://github.com/OUC-VAS/FacePoison.

arXiv

Forensics Adapter: Unleashing CLIP for Generalizable Face Forgery Detection

Xinjie Cui, Yuezun Li, Delong Zhu, Jiaran Zhou, Junyu Dong, Siwei Lyu

We describe Forensics Adapter, an adapter network designed to transform CLIP into an effective and generalizable face forgery detector. Although CLIP is highly versatile, adapting it for face forgery detection is non-trivial as forgery-related knowledge is entangled with a wide range of unrelated knowledge. Existing methods treat CLIP merely as a feature extractor, lacking task-specific adaptation, which limits their effectiveness. To address this, we introduce an adapter to learn face forgery traces -- the blending boundaries unique to forged faces, guided by task-specific objectives. Then we enhance the CLIP visual tokens with a dedicated interaction strategy that communicates knowledge across CLIP and the adapter. Since the adapter is alongside CLIP, its versatility is highly retained, naturally ensuring strong generalizability in face forgery detection. {With only $\textbf{5.7M}$ trainable parameters, our method achieves superior performance across six standard datasets.} Additionally, we describe Forensics Adapter++, an extended method that incorporates textual modality via a newly proposed forgery-aware prompt learning strategy. This extension leads to a further $\textbf{1.3\%}$ performance boost over the original Forensics Adapter. We believe the proposed methods can serve as a baseline for future CLIP-based face forgery detection methods. The code has been released at https://github.com/OUC-VAS/ForensicsAdapter.

arXiv

PatiGonit22K: A Comprehensive Dataset for Solving Complex Bengali MWPs

Swastika Kundu, Azizul Hakim Fayaz, Tashreef Muhammad

Mathematical Word Problems (MWPs) are an important benchmark for evaluating natural language understanding and quantitative reasoning. Despite recent progress in high resource languages, Bengali remains underexplored due to the limited availability of large scale annotated datasets. In this work, we introduce PatiGonit22K, an expanded Bengali MWP dataset containing 22,441 problems, developed by extending the original PatiGonit dataset with a substantially larger collection of complex mathematical problems. The dataset includes both simple and multi operation equations, providing a balanced benchmark for evaluating mathematical reasoning across different difficulty levels. Each problem is carefully translated, annotated, culturally adapted, and verified to ensure linguistic consistency and mathematical correctness. By increasing both the scale and complexity of Bengali MWPs, PatiGonit22K provides a more comprehensive resource for future research on mathematical reasoning and educational NLP applications in low resource languages.

arXiv

Language-Routed RAG and Direct Option Scoring for Multilingual Financial QA: DS@GT at FinMMEval

Justice Ayela, Kabir Sahni

We present DS@GT's submission to FinMMEval 2026 Task 1, a multilingual financial exam question answering benchmark spanning English, Spanish, Greek, Chinese, and Hindi. Financial certification exams such as the CFA, EFPA, and CPA demand structured domain reasoning that standard NLP benchmarks do not capture, and this challenge compounds across languages where retrieval and representation infrastructure is underdeveloped. We build a retrieval-augmented pipeline on LangGraph that detects query language and retrieves semantically relevant exemplars from a 30,209-entry multilingual knowledge base using BGE-M3 embeddings and FAISS indexing. The system then scores answers via Retrieval-Augmented Direct Scoring (RADS), reading next-token log-probabilities over candidate option letters rather than generating free-form output. For low-resource languages, we fuse per-language and cross-lingual retrieval indices using weighted Reciprocal Rank Fusion. Model selection is language-routed: Qwen3-14B for Arabic, Chinese, and Hindi; Qwen2.5-14B for English; and Llama-3.1-8B for Greek, a routing derived from empirical ablations that reveal substantial language-asymmetric performance gaps. Notably, chain-of-thought prompting significantly degrades Greek accuracy (90.7% to 20.9%), and enabling Qwen3's default thinking mode collapses Arabic RADS performance to near-chance levels. Our results indicate that effective multilingual financial reasoning requires language-aware retrieval, model routing, and deliberate scoring strategy selection.

arXiv

Frustratingly Simple Black-Box Adaptation of Language Models via Logit Bias

Ofek I. Cohen, Lior Shani, Aviv Rosenberg, Ankur Samanta, Tal Wagner, Yonathan Efroni

Many organizations aim to adapt language models for internal use, both to improve performance on domain-specific tasks and to address privacy concerns around sensitive data. However, such adaptation remains non-trivial: it often requires operationally challenging fine-tuning of open-source models or ad hoc prompt optimization. We study a minimal alternative based on a simple API-level control: allowing users to bias the model's logits with a user-defined vector. We develop a black-box method for learning a single context-independent logit-bias vector, added at every decoding step, without modifying model weights or requiring gradients. Starting from a KL-regularized reinforcement learning (RL) objective, we characterize when such a fixed logit-bias vector can approximate the optimal prefix-dependent correction and derive a closed-form inverse-propensity estimator from rollouts, rewards, and token probabilities. Empirically, this simple decoding-time intervention improves over base models on mathematical and reasoning benchmarks while using far fewer trainable parameters than conventional fine-tuning. Our results suggest that learned logit bias is a lightweight mechanism for adapting language models under minimal access requirements.

arXiv

The Best Programming Language for Tokenmaxxing: An Investigation of Coding Agent Behavior Across Programming Languages

Zixuan Wu, Carolyn Jane Anderson, Arjun Guha

Although coding agents are now very effective in a variety of programming languages, this paper first shows that the cost (in tokens) can very significantly by programming language. We evaluate five recent models on programming problems in Python, Java, Rust, and OCaml. We carefully control for problem difficulty, and show that there can be stark variation in token consumption that is consistent across models. To understand why, we analyze both the structure and content of agent trajectories. First, we re-execute every intermediate solution and abstract each trajectory as a sequence of test-outcome vectors, then label the work between successive solutions. This reveals agents repeatedly producing noncompiling solutions in unfamiliar languages and revising solutions that already pass. Second, we analyze trajectory text, finding that agents plan solutions in code comments, distrust the provided tests in favor of inputs they invent, and sidestep unfamiliar target languages by prototyping in Python. Our results show that by-language token efficiency is a metric that should be considered when benchmarking and developing multilingual agents, and, for the tokenmaxxer, a guide to the most expensive language to work in.

Keep browsing

More papers from top AI conferences

30+ top conferences, by conference, year, and topic.