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

MetaSyn: A Benchmark for LLM Agents on Meta-Analysis Articles from Nature Portfolio

Anzhe Xie, Weihang Su, Yujia Zhou, Yiqun Liu, Min Zhang, Qingyao Ai

Systematic review and meta-analysis is an important method for scientific research. It comprehensively studies target research questions by combining evidence from multiple independent studies following a standard and rigid protocol (i.e., PI/ECO). This is valuable for facilitating reliable and sustainable research across multiple domains, including but not limited to physics, chemistry, psychology, and medical science. Traditional meta-analysis is both mind-intensive and labor-intensive, as it requires professionals to search, screen, and synthesize tens of thousands of studies, which usually takes months or even years. Recent advances in AI, particularly LLM agents, have the potential to significantly automate this process, but whether they can conduct reliable meta-analysis is still unknown. To this end, we introduce MetaSyn, a dataset and evaluation protocol built from 422 expert-curated meta-analyses manually selected from more than 34,000 published articles in Nature Portfolio journals. We collected the research questions with structured eligibility criteria, the studies included by the original reviewers, and a shared PubMed-anchored corpus containing both eligible studies and plausible but ineligible distractors to build a comprehensive benchmark for meta-analysis. Our experiments on multiple LLMs and baseline methods show that existing AI systems are far from perfect. We also conducted stage-wise evaluation and analysis to shed light on why existing AI systems fall short on meta-analysis.

arXiv

Artificial Intelligence for Mathematical Reasoning: An Integrated Survey of Language Models, Neuro-symbolic Systems, and Verified Discovery

Syed Rifat Raiyan, Mohsinul Kabir, Hasan Mahmud, Md Kamrul Hasan, Sophia Ananiadou

Mathematical reasoning has long served as a stringent test of machine intelligence; over the past decade, it has moved from a niche problem within NLP to one of the most consequential AI frontiers. This survey provides a unified account of the field's evolution, from early rule-based math word problem (MWP) solvers and template-driven geometry systems, through neural expression generation and LLM prompting, to contemporary reasoning models, multi-agent systems, neuro-symbolic theorem provers, and verified discovery workflows. We organize the landscape along four axes: (i) informal reasoning over text and diagrams, spanning MWP solving, multimodal geometry, and VLMs; (ii) formal reasoning in proof assistants, including autoformalization, tactic prediction, compiler-guided repair, and proof search; (iii) mathematical discovery, where systems propose constructions, improve bounds, or assist attacks on open problems; and (iv) the inference and training-time techniques, including CoT prompting, tool use, process reward models, and RLVR, that increasingly connect generation with verification. We catalog major benchmarks across grade-school arithmetic, competition mathematics, geometry, formal proving, multimodal and multilingual reasoning, and expert evaluation, and we examine benchmark saturation, contamination, reporting mismatches, and the distinction between pass@1, majority voting, and verifier-assisted pass@$k$. We critically assess failure modes: brittleness under perturbation, reward hacking, multimodal grounding failures, fragile formalization, and the energy cost of reasoning-scale inference. Drawing on recent perspectives from working mathematicians, we identify future directions centered on verified-discovery workflows, reasoning efficiency, and infrastructure to make AI-assisted formalization broadly usable. Companion materials: https://github.com/Starscream-11813/awesome-AI4Math.

arXiv

StreamProfileBench: A Benchmark for Fine-Grained User Profile Inference in Real-World Streaming Scenarios

Sizhe Wang, Feiyu Duan, Juelin Wang, Liwen Zhang, Zhongyu Wei

Large Language Models (LLMs) have reshaped user profiling, yet current evaluations mainly focus on static data snapshots. This paradigm overlooks the reality of personalized systems, where User-Generated Content (UGC) arrives continuously and fine-grained profiles evolve rapidly. To bridge this gap, we introduce StreamProfileBench, a large-scale benchmark for fine-grained streaming user profiling. We formalize streaming user profiling as a continuous state maintenance task and curate a highly authentic dataset comprising over 120,000 UGC posts from 7,000+ real users across five diverse platforms. By leveraging the temporal correlation of user interests, we further propose a novel, annotation-free evaluation framework. Extensive experiments across 14 leading LLMs reveal that continuous profile updating remains an open challenge. Models exhibit a systemic conservative bias, over-retaining past interests while failing to recognize interest decay. Ablation experiments further validate the practical utility and necessity of the streaming paradigm. Data and code are hosted in https://github.com/WaterWang-001/StreamProfileBench.

arXiv

Repeated-Token Counting Reveals a Dissociation Between Representations and Outputs

Sohan Venkatesh

Large language models fail at counting how many times a word repeats in a list, even though they perform well on far harder reasoning tasks. These failures are commonly attributed to limitations in internal count tracking. We show this attribution is wrong. Linear probes on the residual stream decode the correct count with near-perfect accuracy at every post-embedding layer and they do so even at the exact layers where the wrong answer crystallizes in the output. Attention patterns show no evidence of collapse over repeated tokens and tokenization artifacts account for none of the failure. Instead, a multi-layer perceptron (MLP) block at roughly 85--93\% network depth overwrites the correctly-encoded count with a fixed wrong answer. Ablating this block changes the wrong output and establishes it as causally responsible for the failure. The block fires on the space-separated repeated-word format and is absent for repeated digit-tokens. The pattern holds across Llama-3.2 (1B and 3B) and Qwen2.5 (1.5B, 3B and 7B) at consistent relative depth. The count is represented correctly and a specific computation prevents it from reaching the output, so representation failures and routing failures require different interventions.

arXiv

Procedural Knowledge at Scale Improves Reasoning

Di Wu, Devendra Singh Sachan, Wen-tau Yih, Mingda Chen

Test-time scaling has emerged as an effective way to improve language models on challenging reasoning tasks. However, most existing methods treat each problem in isolation and do not systematically reuse knowledge from prior reasoning trajectories. In particular, they underutilize procedural knowledge: how to reframe a problem, choose an approach, and verify or backtrack when needed. We introduce \textbf{Reasoning Memory}, a retrieval-augmented generation (RAG) framework for reasoning models that explicitly retrieves and reuses procedural knowledge at scale. Starting from existing corpora of step-by-step reasoning trajectories, we decompose each trajectory into self-contained subquestion-subroutine pairs, yielding a datastore of 32 million compact procedural knowledge entries. At inference time, a lightweight in-thought prompt lets the model verbalize the core subquestion, retrieve relevant subroutines within its reasoning trace, and reason under diverse retrieved subroutines as implicit procedural priors. Across six math, science, and coding benchmarks, Reasoning Memory consistently outperforms RAG with document, trajectory, and template knowledge, as well as a compute-matched test-time scaling baseline. With a higher inference budget, averaged across models and tasks, it improves over no retrieval by 19.5% and over the compute-matched baseline by 8.9%. Ablation studies show that these gains come from two key factors: the broad procedural coverage of the source trajectories and our decomposition and retrieval design, which together enable effective extraction and reuse of procedural knowledge. Our experiment code is available at https://github.com/facebookresearch/reasoning-memory.

arXiv

An Efficient and Effective Evaluator for Text2SQL Models on Unseen and Unlabeled Data

Trinh Pham, Thanh Tam Nguyen, Viet Huynh, Hongzhi Yin, Quoc Viet Hung Nguyen

Recent advances in large language models have strengthened Text2SQL systems that translate natural language questions into database queries. A persistent deployment challenge is to assess a newly trained Text2SQL system on an unseen and unlabeled dataset when no verified answers are available. This situation arises frequently because database content and structure evolve, privacy policies slow manual review, and carefully written SQL labels are costly and time-consuming. Without timely evaluation, organizations cannot approve releases or detect failures early. FusionSQL addresses this gap by working with any Text2SQL models and estimating accuracy without reference labels, allowing teams to measure quality on unseen and unlabeled datasets. It analyzes patterns in the system's own outputs to characterize how the target dataset differs from the material used during training. FusionSQL supports pre-release checks, continuous monitoring of new databases, and detection of quality decline. Experiments across diverse application settings and question types show that FusionSQL closely follows actual accuracy and reliably signals emerging issues. Our code is available at https://github.com/phkhanhtrinh23/FusionSQL.

arXiv

From "Help" to Helpful: A Hierarchical Assessment of LLMs in Mental e-Health Applications

Philipp Steigerwald, Jens Albrecht

Psychosocial online counselling frequently encounters generic subject lines that impede efficient case prioritisation. This study evaluates eleven large language models generating six-word subject lines for German counselling emails through hierarchical assessment - first categorising outputs, then ranking within categories to enable manageable evaluation. Nine assessors (counselling professionals and AI systems) enable analysis via Krippendorff's $α$, Spearman's $ρ$, Pearson's $r$ and Kendall's $τ$. Results reveal performance trade-offs between proprietary services and privacy-preserving open-source alternatives, with German fine-tuning consistently improving performance. The study addresses critical ethical considerations for mental health AI deployment including privacy, bias and accountability.

arXiv

RM-Distiller: Exploiting Generative LLM for Reward Model Distillation

Hongli Zhou, Hui Huang, Wei Liu, Chenglong Wang, Xingyuan Bu, Lvyuan Han, Fuhai Song, Muyun Yang, Wenhao Jiang, Hailong Cao, Tiejun Zhao

Reward models (RMs) play a pivotal role in aligning large language models (LLMs) with human preferences. Due to the difficulty of obtaining high-quality human preference annotations, distilling preferences from generative LLMs has emerged as a standard practice. However, existing approaches predominantly treat teacher models as simple binary annotators, failing to fully exploit the rich knowledge and capabilities for RM distillation. To address this, we propose RM-Distiller, a framework designed to systematically exploit the multifaceted capabilities of teacher LLMs: (1) Refinement capability, which synthesizes highly correlated response pairs to create fine-grained and contrastive signals. (2) Scoring capability, which guides the RM in capturing precise preference strength via a margin-aware optimization objective. (3) Generation capability, which incorporates the teacher's generative distribution to regularize the RM to preserve its fundamental linguistic knowledge. Extensive experiments demonstrate that RM-Distiller significantly outperforms traditional distillation methods both on RM benchmarks and reinforcement learning-based alignment, proving that exploiting multifaceted teacher capabilities is critical for effective reward modeling. To the best of our knowledge, this is the first systematic research on RM distillation from generative LLMs.

arXiv

CSV-Decode: Certifiable Sub-Vocabulary Decoding for Efficient Large Language Model Inference

Dong Liu, Shu Wang, Yanxuan Yu, Haisheng Wang, Ben Lengerich

Large language models face significant computational bottlenecks during inference due to the expensive output layer computation over large vocabularies. We present CSV-Decode, a novel approach that uses geometric upper bounds to construct small sub-vocabularies for each decoding step, enabling efficient sparse computation while maintaining dual correctness guarantees: exact top-$k$ certification and $\varepsilon$-certified softmax approximations. Our method clusters vocabulary embeddings offline and uses centroid-plus-radius bounds to identify which tokens can be safely omitted from computation. We provide a complete system implementation with sparse GEMV kernels, multi-GPU sharding, and CUDA Graph optimization. Experimental results demonstrate significant speedup over full vocabulary decoding while maintaining distributional guarantees and low fallback rates. Our code implementation available at \href{https://github.com/FastLM/CSV-Decode}{https://github.com/FastLM/CSV-Decode}.

arXiv

OS-Sentinel: Towards Safety-Enhanced Mobile GUI Agents via Hybrid Validation in Realistic Workflows

Qiushi Sun, Mukai Li, Zhoumianze Liu, Zhihui Xie, Fangzhi Xu, Zhangyue Yin, Kanzhi Cheng, Zehao Li, Zichen Ding, Qi Liu, Zhiyong Wu, Zhuosheng Zhang, Ben Kao, Lingpeng Kong

Computer-using agents powered by Vision-Language Models (VLMs) have demonstrated human-like capabilities in operating digital environments like mobile platforms. While these agents hold great promise for advancing digital automation, their potential for unsafe operations, such as system compromise and privacy leakage, is raising significant concerns. Detecting these safety concerns across the vast and complex operational space of mobile environments presents a formidable challenge that remains critically underexplored. To establish a foundation for mobile agent safety research, we introduce MobileRisk-Live, a dynamic sandbox environment accompanied by a safety detection benchmark comprising realistic trajectories with fine-grained annotations. Built upon this, we propose OS-Sentinel, a novel hybrid safety detection framework that synergistically combines a Formal Verifier for detecting explicit system-level violations with a VLM-based Contextual Judge for assessing contextual risks and agent actions. Experiments show that OS-Sentinel achieves 10%-30% improvements over existing approaches across multiple metrics. Further analysis provides critical insights that foster the development of safer and more reliable autonomous mobile agents. Our code and data are available at https://qiushisun.github.io/OS-Sentinel-Home/.

arXiv

LuxInstruct: A Cross-Lingual Instruction Tuning Dataset For Luxembourgish

Fred Philippy, Laura Bernardy, Siwen Guo, Jacques Klein, Tegawendé F. Bissyandé

Instruction tuning has become a key technique for enhancing the performance of large language models, enabling them to better follow human prompts. However, low-resource languages such as Luxembourgish face severe limitations due to the lack of high-quality instruction datasets. Traditional reliance on machine translation often introduces semantic misalignment and cultural inaccuracies. In this work, we address these challenges by creating a cross-lingual instruction tuning dataset for Luxembourgish, without resorting to machine-generated translations into it. Instead, by leveraging aligned data from English, French, and German, we build a high-quality dataset that preserves linguistic and cultural nuances. We provide evidence that cross-lingual instruction tuning not only improves representational alignment across languages but also the model's generative capabilities in Luxembourgish. This highlights how cross-lingual data curation can avoid the common pitfalls of machine-translated data and directly benefit low-resource language development.

arXiv

The Few-shot Dilemma: Over-prompting Large Language Models

Yongjian Tang, Doruk Tuncel, Christian Koerner, Thomas Runkler

Over-prompting, a phenomenon where excessive examples in prompts lead to diminished performance in Large Language Models (LLMs), challenges the conventional wisdom about in-context few-shot learning. To investigate this few-shot dilemma, we outline a prompting framework that leverages three standard few-shot selection methods - random sampling, semantic embedding, and TF-IDF vectors - and evaluate these methods across multiple LLMs, including GPT-4o, GPT-3.5-turbo, DeepSeek-V3, Gemma-3, LLaMA-3.1, LLaMA-3.2, and Mistral. Our experimental results reveal that incorporating excessive domain-specific examples into prompts can paradoxically degrade performance in certain LLMs, which contradicts the prior empirical conclusion that more relevant few-shot examples universally benefit LLMs. Given the trend of LLM-assisted software engineering and requirement analysis, we experiment with two real-world software requirement classification datasets. By gradually increasing the number of TF-IDF-selected and stratified few-shot examples, we identify their optimal quantity for each LLM. This combined approach achieves superior performance with fewer examples, avoiding the over-prompting problem, thus surpassing the state-of-the-art by 1% in classifying functional and non-functional requirements.

arXiv

Reading Between the Signs: Predicting Future Suicidal Ideation from Adolescent Social Media Texts

Paul Blum, Enrico Liscio, Ruixuan Zhang, Caroline Figueroa, Pradeep K. Murukannaiah

Suicide is a leading cause of death, yet predicting it remains a significant challenge. Risk factors such as depression or substance use are commonly used for prediction, but their predictive performance is often only slightly better than chance. Additionally, many cases go undetected due to a lack of contact with mental health services. Social media, however, offers a unique opportunity, as people often share their thoughts and struggles online in real time. In this work, we propose a novel task and method for early identification: predicting suicidal ideation and behavior (SIB) before a user ever expresses it on an online forum. We introduce Early-SIB, a transformer-based model that sequentially processes the posts a user writes and engages with to predict whether they will write a SIB post. Our model achieves a balanced accuracy of 0.73 in predicting future SIB on a Dutch youth forum, demonstrating that such tools can offer a meaningful addition to traditional methods. Finally, we use Shapley Additive Explanations to make the model's predictions more interpretable.

arXiv

CodeEvo: Interaction-Driven Synthesis of Code-centric Data through Hybrid and Iterative Feedback

Qiushi Sun, Jinyang Gong, Lei Li, Qipeng Guo, Fei Yuan

Acquiring high-quality instruction-code pairs is essential for training Large Language Models for code generation. While automated synthesis has emerged as an alternative to expensive manual curation, current approaches often rely on rigid heuristics, yielding data that is ungrounded or lacks logical complexity. We propose CodeEvo, a dual-agent architecture comprising a Coder for iterative solution synthesis and a Reviewer to orchestrate the generation trajectory. To transcend the limitations of existing heuristics, the Reviewer formulates a Schema to systematically architect logic and complexity through an interleaved synthesis of instructions and code. This process is further reinforced by a hybrid verification protocol synergizing deterministic compiler feedback with semantic evaluation. Under this framework, we construct CodeEvo-100K, a large-scale dataset of instruction-code pairs with stepped difficulty levels. Extensive experiments demonstrate that models fine-tuned on CodeEvo data consistently outperform established baselines across code generation benchmarks. In-depth analyses further provide insights into effective code-centric data synthesis. Code and data are available at https://github.com/QiushiSun/CodeEvo.

arXiv

Measuring Negative Campaigning across Languages with Large Language Models: A Study of 18 Million Tweets in 19 Countries

Victor Hartman, Petter Törnberg

Negative campaigning is a defining feature of electoral competition, yet comparative research on its drivers has remained limited by the high cost and limited scalability of existing classification methods. This study makes two key contributions. First, it evaluates zero-shot large language models (LLMs) as a scalable method for cross-lingual classification of negative campaigning. Using benchmark datasets in ten languages, we show that LLM classifications closely match native-speaker human annotations while outperforming conventional supervised models. Second, we leverage this approach to conduct, to our knowledge, the largest cross-national study of negative campaigning to date, analyzing 18 million tweets posted by parliamentarians in 19 European countries between 2017 and 2022. Building on a strategic incentives framework, we argue that governing and coalition-oriented parties face stronger reputational constraints against negative campaigning, whereas opposition and outsider parties face weaker constraints. We further expect parties located away from the ideological center, especially on the radical right, to rely more heavily on confrontational rhetoric. The results support these expectations: cabinet parties are less negative, while ideologically non-centrist parties--most notably radical right parties--are substantially more negative. The study thus provides new comparative evidence on the party-level foundations of campaign negativity while demonstrating how LLMs can transform the scale, consistency, and replicability of research on political discourse across languages and institutional contexts.

arXiv

PD$^3$: A Project Duplication Detection Framework via Adapted Multi-Agent Debate

Dezheng Bao, Yueci Yang, Chutian Yu, Xin Chen, Zeguo Fei, Xiang Yuan, Lijun Zhang, Jiangqian Huang, Zhengxuan Jiang, Daoze Zhang, Junru Chen, Yang Yang

Project duplication detection is critical for project quality assessment because it helps avoid investment in repeated proposals. Existing methods usually cast it as ranking and rely on surface matching or direct large language models judging, often missing practical needs in set-level reference selection. We recast the task as many-to-many reference set selection, which requires broad candidate information and fair decomposed comparison under context limits. We propose PD$^3$, a framework for Project Duplication Detection via adapted multi-agent Debate. PD$^3$ combines local multi-agent debate with global round-robin scheduling to retrieve the relevant project set. Theoretically, this scheduler guarantees fair comparison through balanced exposure and comparison context. PD$^3$ also produces quantitative duplication scores and qualitative overlap feedback. On 800+ real-world power projects, PD$^3$ outperforms the strongest baselines by 4.05% in relevant reference selection and 9.77% in duplication score generation. We deploy Review Dingdang, an online platform, which has helped save $13.44 million across 442 new projects.

arXiv

Identifiability-Aware Source Apportionment in City-Scale Advection-Diffusion Systems

Ankit Bhardwaj, Lakshminarayanan Subramanian

Source apportionment from sparse urban air-quality sensors is an inverse problem limited by sensor placement, wind-driven transport, background variation, and noise. Known or proxy emission inventories make attribution meaningful by restricting the unknown source field to a finite set of candidate groups, but do not guarantee those groups are distinguishable from the observations. We represent time-varying source activity with a low-dimensional nonnegative temporal basis and formulate inventory-based apportionment as a wind-conditioned lagged inverse problem in which each source--basis coefficient produces a sensor-time fingerprint. After projecting out a separate low-dimensional background space, the relevant object is the projected lagged response matrix $\widetilde H_Φ$: exact identifiability at the chosen basis resolution requires its full column rank, while noise-robust attribution is controlled by its singular values, coefficient visibility, background absorption, pairwise coherence, and ray distance. We propose an identifiability-aware apportionment (IASA) framework that estimates nonnegative source--basis coefficients, reconstructs activity trajectories, and reports uncertainty and conservative grouping recommendations for indistinguishable sources. We instantiate it on a New Delhi platform built from government PM$_{2.5}$ and wind records, regulatory sensor locations, and four proxy source groups, and define controlled and observed evaluations of recovery, ambiguity, wind diversity, background stress, transport error, inventory robustness, and residual adequacy. IASA reports the attribution resolution defensible under the declared inventories, transport, background, lag, and noise rather than the most detailed possible vector.

arXiv

Not All EEG Moments Are Equal: Position-Adaptive Time Scheduling for EEG Generation

boheng liu, Ziyu Li, Chenghua Duan, Qing Li, Xia Wu

Electroencephalography (EEG) generation is essential for alleviating data scarcity and enabling large scale neural modeling in brain computer interface applications. However, existing flow based approaches assume that every channel and every time segment within a sample shares a single global time progression, overlooking the fact that not all EEG moments are equal. To address this overlooked heterogeneity, we propose an adaptive EEG generation framework built on conditional flow matching. The framework introduces Position-Adaptive Time Scheduling, which tracks per position reconstruction error to modulate a position specific time progress within the flow matching trajectory. It further incorporates Factorized Spatio-Temporal Attention and a frequency aligned multi resolution spectral consistency loss to model inter channel dependencies induced by volume conduction and compensate for the power law spectral bias of EEG, thereby improving the quality of generated signals. Extensive experiments on three EEG datasets with distinct acquisition protocols and task semantics show that our framework consistently outperforms the strongest baseline, reducing TS-FID by up to 62.2\% and improving downstream classification accuracy gain by up to 6.77 percentage points. These results suggest that the proposed method represents a promising step toward scalable, high fidelity data augmentation for real world brain computer interface applications.

arXiv

Fracture Risk Prediction in Adults Over 50 Years Old Using DXA and EHR: Comparison of Traditional and Machine Learning Models in Two Large Cohorts

Jiahe Qian, Hao Dai, Kunyu Yu, Hexin Dong, Xing He, Erik A. Imel, Jiang Bian, Yifan Peng, Yi Liu

Accurate fracture risk prediction is important for osteoporosis management, but commonly used clinical tools may not fully use information available in electronic health records (EHRs) and dual-energy X-ray absorptiometry (DXA) reports. We developed and externally validated time-to-event fracture prediction models among adults aged 50 years or older with clinically obtained DXA reports in 2 US health care systems. The development cohort was derived from NewYork-Presbyterian/Weill Cornell Medical Center and the external validation cohort from the Indiana Network for Patient Care. Predictors included demographics, lifestyle factors, prior fracture, comorbidities, medication exposures, osteoporosis treatment history, and DXA-derived T-scores extracted from radiology reports. The outcome was time from index DXA to first incident fragility fracture identified from structured diagnosis codes. We evaluated penalized Cox regression, random survival forest, gradient-boosting survival, and XGBoost survival models using 2 prespecified predictor settings and compared discrimination with clinically reported FRAX major osteoporotic fracture probabilities. The development cohort included 11,510 adults, of whom 858 sustained incident fragility fractures; the external validation cohort included 1,932 adults, of whom 180 sustained fractures. In internal validation, the expanded Cox model achieved a mean Harrell C-index of 0.779, compared with 0.653 for FRAX. In external validation, the corresponding Cox model achieved a Harrell C-index of 0.714, compared with 0.590 for FRAX; gradient-boosting survival had the highest external discrimination (0.725). EHR- and DXA-enhanced models showed better discrimination than clinically reported FRAX scores in this DXA-tested population, but calibration assessment, prospective evaluation, and implementation workflow assessment are needed before clinical use.

arXiv

LARA: Lightweight Adapters in the Residual Stream for Composable Adaptation and Alignment

Pascal Ekin, Hyosun Choi, Wei Jie

We present LARA (Lightweight Additive Residual Adaptation), a method for efficient adaptation that operates in the residual stream of a frozen model rather than in its weights. Where LoRA adds an update of low rank to weight matrices, LARA reads the hidden state at a small set of layers and adds a correction of low rank back to the residual stream, leaving all base weights untouched. On a code fine-tuning task and on preference optimization (DPO), LARA matches LoRA at equal parameter counts. Because adaptation is a frozen base plus a residual, LARA exposes a scale γ, applied at inference, that interpolates smoothly between base and adapted behavior, a form of graded control that adaptation in weight space does not offer. Finally, because each behavior is a small residual module over a shared frozen base, many behaviors can be held resident at once and routed automatically per token. We place seven behaviors, six fine-tuned and one optimized for preference, on one frozen 1.5B model for roughly 33 MB of overhead, against one full model for each behavior. Because the base is untouched, behaviors are trained separately and selected per token rather than loaded on demand, which suits hosting many behaviors, and adding new ones, on a single model on a device.

Keep browsing

More papers from top AI conferences

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