Singapore

Large Language Model (LLM) agent systems have advanced rapidly, driven by their strong generalization in zero-shot settings. To further enhance reasoning and accuracy on complex tasks, Multi-Agent Debate (MAD) has emerged as a promising framework that engages multiple LLM agents in structured debates to encourage diverse reasoning. However, triggering MAD for every input instance is inefficient, as it incurs substantial computational (token) cost and may even degrade accuracy by overturning correct single-agent answers. To address these limitations, we propose intelligent Multi-Agent Debate (iMAD), a token-efficient framework that selectively triggers MAD only when it is likely to be beneficial (i.e., correcting an initially wrong answer) in the zero-shot setting. To achieve this goal, iMAD learns generalizable model behaviors to make accurate debate decisions in the zero-shot setting. Specifically, it first prompts a single agent to produce a structured self-critique response, from which we extract over 40 interpretable linguistic and semantic features capturing hesitation cues. A lightweight classifier, based on Multi-Layer Perceptron and trained using our proposed FocusCal loss, then determines whether to trigger MAD, enabling robust zero-shot decisions without dataset-specific tuning. We evaluate iMAD on six (visual) question answering datasets against five competitive baselines. iMAD significantly reduces token usage (by up to 92%) while also improving final answer accuracy (by up to 13.5%).

AAAI 2026

iMAD: Intelligent Multi-Agent Debate for Efficient and Accurate LLM Inference

mas: other foundations of multi agent systems

ml: efficient ml / green ai

argumentation & negotiation

mas: agreement

### Welcome to 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics

Welcome to the 2025 meeting of the Nations of the Americas Chapter of the Association for Computational Linguistics! I am proud to help organize the first NAACL conference to carry the new name of our organization, one that emphasizes inclusion for all of the Americas. I am also pleased to welcome you to Albuquerque, New Mexico, a state whose unique blend of cultural influences will make for an excellent backdrop for NAACL 2025, especially with this year’s special theme on NLP in a Multicultural World. 
**[Continue reading...](https://drive.google.com/file/d/1jX-qGhqVSZZCIrAnJaz798pflu5Irrdn/view?usp=sharing)**

*- Colin Cherry, Google, NAACL 2025 General Chair* 

[![](https://assets.underline.io/markdown_image/1/image/b087f8a4dc5816d6e1a6514e59c59ac3.png)](https://drive.google.com/file/d/1T96GzPqObXrMln2BMByCSXTSizjTg69P/view?usp=sharing)

You need to log in with the email address you registered with. Access credentials have been sent to your email. 

Please be sure to check your spam and other email folders if you do not see an email confirmation right away.

Please log in to explore this event.

To access NAACL 2025 event page you are required to register. Please follow [**this link**](https://2025.naacl.org/registration/registration/) to register. Access will depend on your registration type.

Please register!

NAACL 2025

Welcome to 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics

technical paper

We are pleased to announce the Fortieth AAAI Conference on Artificial Intelligence (AAAI-26), which will be held in Singapore EXPO from January 20 to January 27, 2026.

The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines. AAAI-26 will feature technical paper presentations, special tracks, invited speakers, workshops, tutorials, poster sessions, senior member presentations, competitions, and exhibit programs, and a range of other activities to be announced.

To access this event page, you need to log in with the **email address you registered with**. Access credentials will be sent to your email from Underline - subject line "Welcome to AAAI 2026". Please be sure to check your spam email folder if you do not see an email confirmation right away.

Please log in

To access this event page, you are required to register.
Please complete your registration to continue.

We recommend reading [**the registration information**](https://aaai.org/conference/aaai/aaai-26/registration/) first.

**Online Registration Form**: https://aaai.getregistered.net/conference-2026 

Registration Required

We are pleased to announce the Fortieth AAAI Conference on Artificial Intelligence (AAAI-26), which will be held in Singapore EXPO from January 20 to January 27, 2026.

We investigate whether large language models can discover and analyze U.S. tax-minimization strategies. This real-world domain challenges even seasoned human experts, and progress can reduce tax revenue lost from well-advised, wealthy taxpayers. We evaluate the most advanced LLMs on their ability to (1) interpret and verify tax strategies, (2) fill in gaps in partially specified strategies, and (3) generate complete, end-to-end strategies from scratch. This domain should be of particular interest to the LLM reasoning community: unlike synthetic challenge problems or scientific reasoning tasks, U.S. tax law involves navigating hundreds of thousands of pages of statutes, case law, and administrative guidance, all of which is updated regularly. Notably, LLM-based reasoning identified an entirely novel tax strategy, highlighting these models' potential for revolutionizing tax agencies' fight against tax abuse.

Can LLMs Identify Tax Abuse?

Conditional molecular generation, aiming to generate 2D & 3D molecules that satisfy given properties, has achieved remarkable progress, thanks to the advances in deep generative models such as graph diffusion. However, existing methods generally assume that the given conditions for training and testing are consistent, failing to handle the realistic challenge when there exist distribution shifts between training and testing conditions. Invariant learning is a mainstream paradigm for addressing distribution shifts, but fusing invariant learning principles with conditional molecular generation faces three core challenges: (1) existing invariant learning methods focus on discriminative tasks and cannot be directly adapted to molecule generative tasks; (2) how to distinguish between invariant subgraph and variant subgraph of a molecule graph, which is treated as an integrated input; (3) how to fuse invariant subgraphs, variant subgraphs, and property conditions for effective generation. To tackle these challenges, we propose $\underline{I}$nvariant $\underline{C}$onditional $\underline{MOL}$ecular generation (IC-MOL), a framework that combines invariant learning with graph diffusion to improve the generalization ability of conditional molecular generation under distribution shifts. Specifically, we first disentangle molecular graphs into invariant and variant subgraphs while maintaining SE(3) equivariance, an important inductive bias for molecular generation. On this basis, we further design a two-phase graph diffusion generation model. In the first phase, we generate an invariant molecular consistent with the target property. In the second phase, we propose a cross-attention mechanism to fuse variant subgraph representations and property conditions to guide the generation of complete molecules while maintaining property alignment. Extensive experiments on the benchmark dataset show that IC-MOL consistently outperforms state-of-the-art baselines across six property conditions under distribution shifts.

Invariant Conditional Molecular Generation Under Distribution Shift

Many sequential decision-making problems can be formulated as shortest-path problems, where the objective is to reach a goal state from a given starting state. Heuristic search is a standard approach for solving such problems, relying on a heuristic function to estimate the cost to the goal from any given state. Recent approaches leverage reinforcement learning to learn heuristics by applying deep approximate value iteration. These methods typically rely on single-step Bellman updates, where the heuristic of a state is updated based on its best neighbor and the corresponding edge cost. This work proposes a generalized approach that enhances both state sampling and heuristic updates by performing limited-horizon searches and updating each state's heuristic based on the shortest path to the search frontier, incorporating both edge costs and the heuristic values of frontier states.

Beyond Single-Step Updates: Reinforcement Learning of Heuristics with Limited-Horizon Search

This work introduces Text-based Aerial-Ground Person Retrieval (TAG-PR), which aims to retrieve person images from heterogeneous aerial and ground views with textual descriptions. 
Unlike traditional Text-based Person Retrieval (T-PR), which focuses solely on ground-view images, TAG-PR introduces greater practical significance and presents unique challenges due to the large viewpoint discrepancy across images. 
To support this task, we contribute: (1) TAG-PEDES dataset, constructed from public benchmarks with automatically generated textual descriptions, enhanced by a diversified text generation paradigm to ensure robustness under view heterogeneity; and (2) TAG-CLIP, a novel retrieval framework that addresses view heterogeneity through a hierarchically-routed mixture of experts module to learn view-specific and view-agnostic features and a viewpoint decoupling strategy to decouple view-specific features for better cross-modal alignment.
We evaluate the effectiveness of TAG-CLIP on both the proposed TAG-PEDES dataset and existing T-PR benchmarks. The dataset and code will be publicly available.

Text-based Aerial-Ground Person Retrieval

The high memory demands of the Key-Value (KV) Cache during the inference of Large Language Models (LLMs) severely restrict their deployment in resource-constrained platforms. Quantization can effectively alleviate the memory pressure caused by KV Cache. However, existing methods either rely on static one-size-fits-all precision allocation or fail to dynamically prioritize critical KV in long-context tasks, forcing memory-accuracy-throughput tradeoffs. In this work, we propose a novel mixed-precision quantization method for KV Cache named KVmix. KVmix leverages gradient-based importance analysis to evaluate how individual Key and Value projection matrices affect the model loss, enabling layer-specific bit-width allocation for mix-precision quantization. It dynamically prioritizes higher precision for important layers while aggressively quantizing less influential ones, achieving a tunable balance between accuracy and efficiency. KVmix introduces a dynamic long-context optimization strategy that adaptively keeps full-precision KV pairs for recent pivotal tokens and compresses older ones, achieving high-quality sequence generation with low memory usage. Additionally, KVmix provides efficient low-bit quantization and CUDA kernels to optimize computational overhead. On LLMs such as Llama and Mistral, KVmix achieves near-lossless inference performance with extremely low quantization configuration (Key 2.19bit Value 2.38bit), while delivering a remarkable 4.9× memory compression and a 5.3× speedup in inference throughput.

KVmix: Gradient-Based Layer Importance-Aware Mixed-Precision Quantization for KV Cache

In warehouse-based e-commerce, accurate category-level warehouse demand prediction is essential to ensure effective inventory management. Existing works mainly explore advanced time series models to capture the temporal dynamics, failing to mine cross-category and cross-city correlations effectively. In this paper, we explore large language models as an opportunity to understand the semantic information of warehouses and categories to enhance prediction. However, it is not trivial due to: i) the inaccurate LLM’s understanding of the category-related and warehouse-related textual input; and ii) the cross-warehouse knowledge extraction. To solve the above challenges, we propose an LLM-guided multi-task graph learning framework, LMGL-WD, for category-level warehouse demand prediction. LMGL-WD includes three components: i) an LLM-guided category series encoding module to represent each category through contextual and series embedding; ii) a cross-warehouse category learning module to adaptively mine the informative knowledge from cross-warehouses to enhance category representation; and iii) a cross-category multi-task learning module to adaptively capture cross-category correlations to further improve prediction. Extensive evaluation results with real-world data collected from one of the largest e-commerce platforms in China demonstrate that LMGL-WD achieves superior performance, e.g., reduces MAPE by up to 31.59%, compared to
state-of-the-art methods.

LMGL-WD: LLM-Guided Multi-Task Graph Learning for Category-Level Warehouse Demand Prediction in E-Commerce

LiDAR data generation has emerged as a promising solution to the high cost and limited scalability of real-world LiDAR sensing. Recent diffusion and rectified flow models have demonstrated strong capabilities in synthesizing realistic 3D point clouds; however, their iterative sampling procedures result in significant inference overhead. To address this, we focus on efficient few-step LiDAR generation for both unconditional and multi-modal conditional settings. Specifically, we propose an adaptive piecewise distillation strategy tailored for rectified flow-based LiDAR generation models, where the teacher model’s flow trajectory is adaptively segmented into consecutive intervals, and the student is trained only at the start of each interval to directly predict the velocity toward its endpoint. By sequentially sampling at the start timestep of each interval, our method enables fast few-step generation. Moreover, instead of uniform partitioning, we introduce an adaptive timestep selection strategy that chooses interval boundaries with minimal initial error, thereby reducing the complexity of distillation. Experimental results show that our method achieves comparable or superior performance to state-of-the-art methods in both unconditional and multi-modal conditional LiDAR generation, using only four sampling steps.

Adaptive Piecewise Distillation for Efﬁcient LiDAR Data Generation

Detecting minimal conflict sets (MCSs) is essential for providing meaningful feedback in knowledge-based configuration. While lazy conflict detection addresses runtime efficiency by predetermining conflict sets offline using a genetic algorithm, it suffers from low conflict coverage, stagnation, and instability. We propose a robust enhancement that integrates multi-conflict extraction and genetic diversity control to overcome these limitations. Our method extends conflict discovery per evaluation and introduces three diversity mechanisms: full population reproduction, weighted genetic operators, and adaptive extinction. Empirical evaluations on five real-world configuration knowledge bases show that our approach recovers over 85% of conflict sets on average, reduces solver calls by up to 73%, and achieves higher result stability. These improvements demonstrate the scalability and reliability of enhanced lazy conflict detection for interactive configuration systems.

Robust Lazy Conflict Detection via Multi-Conflict Extraction and Genetic Diversity Control

Speculative decoding accelerates language model inference by separating generation into fast drafting and parallel verification. Its main limitation is drafter–verifier misalignment, which limits token acceptance and reduces overall effectiveness. While small drafting heads trained from scratch compensate with speed, they struggle when verification dominates latency or when inputs are out of distribution. In contrast, pretrained drafters, though slower, achieve higher acceptance rates thanks to stronger standalone generation capabilities, making them competitive when drafting latency is negligible relative to verification or communication overhead. In this work, we aim to improve the acceptance rates of pretrained drafters by introducing a lightweight dynamic alignment mechanism: a \emph{steering vector} computed from the verifier’s hidden states and injected into the pretrained drafter. Compared to existing offline alignment methods such as distillation, our approach boosts the number of accepted tokens by up to 35\% under standard sampling and 22\% under greedy sampling, all while incurring negligible computational overhead. Importantly, our approach can be retrofitted to existing architectures and pretrained models, enabling rapid adoption.

Steering Pretrained Drafters During Speculative Decoding

In many planning problems there are non-deterministic actions for which the outcome cannot be fully controlled by the planning agent. For critical tasks, we need to ensure that the strategy achieves the goal within a predictable time-frame and/or cost. Thus, we consider an adversarial planning setting and compute optimal policies that optimize the worst-case cost to reach the goal.

In this work, we introduce domain-independent optimal heuristic search algorithms for this adversarial setting. To guide the search, we show how to leverage classical planning heuristics by applying single-outcome determinization. We also generalize dominance techniques, that analyse when a state is as good as another, to the non-deterministic setting and apply them to prune the search space. Our experimental analysis shows that both methods greatly helps to compute optimal policies across multiple domains.

Downloads

Next from AAAI 2026

Can LLMs Identify Tax Abuse?

Similar lecture

S2-MAD: Breaking the Token Barrier to Enhance Multi-Agent Debate Efficiency