Singapore

Large Reasoning Models (LRMs) extend large language models with explicit, multi-step reasoning traces to enhance transparency and performance on complex tasks.
However, these reasoning traces can be redundant or logically inconsistent, becoming a new and hard-to-detect source of hallucination.
Existing hallucination detection methods focus primarily on answer-level uncertainty and often fail to detect hallucinations or logical inconsistencies arising from the model’s reasoning trace.
This oversight is particularly problematic for LRMs, where the explicit thinking trace is not only an important support to the model&#39;s decision-making process but also a key source of potential hallucination. 
To this end, we propose RACE (Reasoning and Answer Consistency Evaluation), a novel framework specifically tailored for hallucination detection in LRMs.
RACE operates by extracting essential reasoning steps and computing four diagnostic signals: inter-sample consistency of reasoning traces, entropy-based answer uncertainty, semantic alignment between reasoning and answers, and internal coherence of reasoning. This joint analysis enables fine-grained hallucination detection even when the final answer appears correct.
Experiments across datasets and different LLMs demonstrate that RACE outperforms existing hallucination detection baselines, offering a robust and generalizable solution for evaluating LRMs.

AAAI 2026

Joint Evaluation of Answer and Reasoning Consistency for Hallucination Detection in Large Reasoning Models

nlp: fact-checking / misinformation detection (nlp focus)

nlp: (large) language models

and evaluation of nlp models

nlp: interpretability

analysis

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.<br><br>

To access this event page, you need to log in with the **email address you registered with**. <br>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.

Large Reasoning Models (LRMs) extend large language models with explicit, multi-step reasoning traces to enhance transparency and performance on complex tasks.
However, these reasoning traces can be redundant or logically inconsistent, becoming a new and hard-to-detect source of hallucination.
Existing hallucination detection methods focus primarily on answer-level uncertainty and often fail to detect hallucinations or logical inconsistencies arising from the model’s reasoning trace.
This oversight is particularly problematic for LRMs, where the explicit thinking trace is not only an important support to the model's decision-making process but also a key source of potential hallucination. 
To this end, we propose RACE (Reasoning and Answer Consistency Evaluation), a novel framework specifically tailored for hallucination detection in LRMs.
RACE operates by extracting essential reasoning steps and computing four diagnostic signals: inter-sample consistency of reasoning traces, entropy-based answer uncertainty, semantic alignment between reasoning and answers, and internal coherence of reasoning. This joint analysis enables fine-grained hallucination detection even when the final answer appears correct.
Experiments across datasets and different LLMs demonstrate that RACE outperforms existing hallucination detection baselines, offering a robust and generalizable solution for evaluating LRMs.

technical paper

Machine learning models trained on Earth observation data, such as satellite imagery, have demonstrated significant promise in predicting household-level wealth indices, enabling the creation of high-resolution wealth maps that can be leveraged across multiple causal trials while addressing chronic data scarcity in global development research. However, because standard training objectives prioritize overall predictive accuracy, these predictions inherently suffer from shrinkage toward the mean, leading to attenuated estimates of causal treatment effects and limiting their utility in policy evaluations. Existing debiasing methods, such as Prediction-Powered Inference (PPI), can handle this attenuation bias but require additional fresh ground-truth data at the downstream stage of causal inference, which restricts their applicability in data-scarce environments. In this paper, we introduce and evaluate two correction methods—linear calibration correction and Tweedie's correction—that substantially reduce prediction bias without relying on newly collected labeled data. Linear calibration (LCC) corrects bias through a straightforward linear transformation derived from held-out calibration data, whereas Tweedie's correction leverages empirical Bayes principles to directly address shrinkage-induced biases by exploiting score functions derived from evaluating the model's learning patterns. Through analytical exercises and experiments using Demographic and Health Survey (DHS) data, we demonstrate that both proposed methods meet or outperform existing approaches that either require (a) adjustments to training pipelines or (b) additional labeled data, achieving significant reductions in attenuation bias in data-scarce environments. These approaches may represent a promising avenue for improving the reliability of causal inference when direct outcome measures are limited or unavailable, enabling a ``one map, many trials'' paradigm where a single upstream data creation team produces predictions usable by many downstream teams across diverse ML pipelines.

Debiasing Machine Learning Predictions for Causal Inference Without Additional Ground Truth Data: “One Map, Many Trials” in Satellite-Driven Poverty Analysis

Recent learning-based methods have reduced the computational complexity of traditional trajectory similarity computation, but state-of-the-art (SOTA) methods still fail to leverage the comprehensive spectrum of trajectory information for similarity modeling. To tackle this problem, we propose \textbf{RePo}, a novel method that jointly encodes \textbf{Re}gion-wise and \textbf{Po}int-wise features to capture both spatial context and fine-grained moving patterns. For region-wise representation, the GPS trajectories are first mapped to grid sequences, and spatial context are captured by structural features and semantic context enriched by visual features. For point-wise representation, three lightweight expert networks extract local, correlation, and continuous movement patterns from dense GPS sequences. Then, a router network adaptively fuses the learned point-wise features, which are subsequently combined with region-wise features using cross-attention to produce the final trajectory embedding. To train RePo, we adopt a contrastive loss with hard negative samples to provide similarity ranking supervision. Experiment results show that RePo achieves an average accuracy improvement of 22.2\% over SOTA baselines across all evaluation metrics.

Region-Point Joint Representation for Effective Trajectory Similarity Learning

Multimodal Large Language Models struggle to maintain reliable performance under extreme real-world visual degradations, which impede their practical robustness. Current robustness enhancement methods rely on implicit training/adaptation that focuses solely on visual encoder generalization, suffering from limited interpretability and isolated optimization. To overcome these limitations, we propose Robust-R1, a novel framework that explicitly models visual degradations through structured reasoning chains. Our approach integrates: (i) supervised fine-tuning for degradation-aware reasoning foundations, (ii) reward-driven alignment for accurately perceiving degradation parameters, and (iii) dynamic reasoning depth scaling adapted to degradation intensity. To support this methodology, we introduce a novel 11K dataset featuring realistic degradations synthesized across four critical real-world visual processing stages, each annotated with structured chains connecting degradation parameters, perceptual effects, and pristine semantic reasoning. Comprehensive evaluations demonstrate state-of-the-art robustness: Robust-R1 outperforms all general and robust baselines on the real-world degradation benchmark R-Bench, while maintaining superior anti-degradation performance under multi-intensity adversarial degradations on MMBench, MMStar, and RealWorldQA.
We will release our code, demo, and dataset soon.

Robust-R1: Degradation-Aware Reasoning for Robust Visual Understanding

Recently Multimodal Large Language Models (MLLMs) have achieved considerable advancements in vision-language tasks, yet produce potentially harmful or untrustworthy content. Despite substantial work investigating the trustworthiness of language models, MMLMs' capability to act honestly, especially when faced with visually unanswerable questions, remains largely underexplored. This work presents the first systematic assessment of honesty behaviors across various MLLMs. We ground honesty in models' response behaviors to unanswerable visual questions, define four representative types of such questions, and construct MoHoBench, a large-scale MMLM honest benchmark, consisting of 12k+ visual question samples, whose quality is guaranteed by multi-stage filtering and human verification. Using MoHoBench, we benchmarked the honesty of 28 popular MMLMs and conducted a comprehensive analysis. Our findings show that: (1) most models fail to appropriately refuse to answer when necessary, and (2) MMLMs' honesty is not solely a language modeling issue, but is deeply influenced by visual information, necessitating the development of dedicated methods for multimodal honesty alignment. Therefore, we implemented initial alignment methods using supervised and preference learning to improve honesty behavior, providing a foundation for future work on trustworthy MLLMs.

MoHoBench: Assessing Honesty of Multimodal Large Language Models via Unanswerable Visual Questions

Point-of-Interest (POI) recommendation plays a pivotal role in location-based services by guiding users to discover new and relevant places. While graph-based methods have shown promising results, effectively modeling the diversity and dynamics of user preferences remains a key challenge. Addressing this requires richer representations of both POIs and user interests, as well as more adaptive learning strategies.
In this work, we propose TMHKG, a Task-aware Meta-learning framework with a Heterogeneous Knowledge Graph for POI recommendation. To enhance representation learning, TMHKG constructs a dual-view POI knowledge graph that integrates geographical proximity and user-aware category transitions, and models users' evolving interests from sequential visit histories. On top of enriched features, TMHKG adopts a task-aware meta-learning paradigm, treating each user's recommendation task as a separate meta-task. A generalizable recommendation policy is first learned from diverse training tasks and then quickly adapted to each user's unique behavior, enabling highly personalized predictions.
Extensive experiments on two real-world datasets demonstrate that TMHKG consistently outperforms state-of-the-art baselines, highlighting its effectiveness in capturing complex user-POI interactions.

Task-Aware Meta-Learning on Heterogeneous Knowledge Graph for POI Recommendation

Pre-trained Vision-Language Models (VLMs), e.g. CLIP,
have become essential tools in multimodal transfer learn-
ing. However, fine-tuning VLMs in few-shot scenarios poses
significant challenges in balancing task-specific adaptation
and generalization in the obtained model. Meanwhile, cur-
rent researches have predominantly focused on prompt-based
adaptation methods, leaving adapter-based approaches un-
derexplored and revealing notable performance gaps. To ad-
dress these challenges, we introduce a novel Reconstruction-
based Multimodal Adapter (RMAdapter), which leverages a
dual-branch architecture. Unlike conventional single-branch
adapters, RMAdapter consists of: (1) an adaptation branch
that injects task-specific knowledge through parameter-
efficient fine-tuning, and (2) a reconstruction branch that pre-
serves general knowledge by reconstructing latent space fea-
tures back into the original feature space. This design facil-
itates a dynamic balance between general and task-specific
knowledge. Importantly, although RMAdapter introduces an
additional reconstruction branch, it is carefully optimized
to remain lightweight. By computing reconstruction loss lo-
cally at each layer and sharing projection modules, the over-
all computational overhead is kept minimal. A consistency
constraint is also incorporated to better regulate the trade-
off between discriminability and generalization. We compre-
hensively evaluate the effectiveness of RMAdapter on three
representative tasks: generalization to new categories, gen-
eralization to new target datasets, and domain generalization.
Without relying on data augmentation or duplicate prompt de-
signs, our RMAdapter consistently outperforms state-of-the-
art approaches across all evaluation metrics.

RMAdapter: Reconstruction-based Multi-Modal Adapter for Vision-Language Models

Efficiently and accurately determining the symmetry is a crucial step in the structural analysis of crystalline materials. Existing methods usually mindlessly apply deep learning models while ignoring the underlying chemical rules. More importantly, experiments show that they face a serious sub-property confusion SPC problem. To address the above challenges, from a decoupled perspective, we introduce the XRDecoupler framework, a problem-solving arsenal specifically designed to tackle the SPC problem. Imitating the thinking process of chemists, we innovatively incorporate multidimensional crystal symmetry information as superclass guidance to ensure that the model's prediction process aligns with chemical intuition. We further design a hierarchical PXRD pattern learning model and a multi-objective optimization approach to achieve high-quality representation and balanced optimization. Comprehensive evaluations on three mainstream databases (e.g., CCDC, CoREMOF, and InorganicData) demonstrate that XRDecoupler excels in performance, interpretability, and generalization. The code for our method is available in Supplement.

Rethinking Crystal Symmetry Prediction: A Decoupled Perspective

Imputing missing values in spatial-temporal traffic data is essential for intelligent transportation systems. Among advanced imputation methods, score-based diffusion models have demonstrated competitive performance. These models generate data by reversing a noising process, using observed values as conditional guidance. However, existing diffusion models typically apply a uniform guidance scale across both spatial and temporal dimensions, which is inadequate for nodes with high missing data rates. Sparse observations provide insufficient conditional guidance, causing the generative process to drift toward the learned prior distribution rather than closely following the conditional observations, resulting in suboptimal imputation performance.

To address this, we propose FENCE (Spatial-Temporal Feedback Diffusion Guidance), a novel method that adaptively controls guidance scales during imputation. First, FENCE introduces a dynamic feedback mechanism that adjusts the guidance scale based on the posterior likelihood approximations. The guidance scale is increased when generated values diverge from observations and reduced when alignment improves, preventing overcorrection. Second, because alignment to observations varies across nodes and denoising steps, a global guidance scale for all nodes is suboptimal. FENCE computes guidance scales at the cluster level by grouping nodes based on their attention scores, leveraging spatial-temporal correlations to provide more accurate guidance. Experimental results on real-world traffic datasets show that FENCE significantly enhances imputation accuracy.

Spatial-Temporal Feedback Diffusion Guidance for Controlled Traffic Imputation

While large vision-language models (VLMs) demonstrate impressive long-context understanding, their prevalent small branches fails on linguistics-photography alignment for limited window size. We discover that knowledge distillation improve students capability as compelementary to Rotary Position Embeddings (RoPE) on certain windows size (anchored from large models). Building on this insight, we propose LAid, which explicitly targets the transfer of long-range attention mechanisms through two complementary components: (1) a progressive distance-weighted attention matching that dynamically emphasizes longer position differences during training, and (2) a learnable RoPE response gain modulation that selectively amplifies position sensitivity where needed. Extensive experiments across multiple model families demonstrate that LAid-distilled models achieve up to 3.2× longer effective context windows compared to baseline small models, while maintaining or improving performance on standard VL benchmarks. Spectral analysis also suggests that LAid successfully preserves crucial low-frequency attention components that conventional methods fail to transfer. Our work not only provides practical techniques for building more efficient long-context VLMs but also offers theoretical insights into how positional understanding emerges and transfers during distillation.

Towards Long-window Anchoring in Vision-Language Model Distillation

Multimodal Large Language Models are primarily trained and evaluated on aligned image-text pairs, which leaves their ability to detect and resolve real-world inconsistencies largely unexplored. In open-domain applications visual and textual cues often conflict, requiring models to perform structured reasoning beyond surface-level alignment.
We introduce CrossCheck-Bench, a diagnostic benchmark for evaluating contradiction detection in multimodal inputs. The benchmark adopts a hierarchical task framework covering three levels of reasoning complexity and defines seven atomic capabilities essential for resolving cross-modal inconsistencies. CrossCheck-Bench includes 15k question-answer pairs sourced from real-world artifacts with synthetically injected contradictions. The dataset is constructed through a multi-stage annotation pipeline involving more than 450 expert hours to ensure semantic validity and calibrated difficulty across perception, integration, and reasoning.
We evaluate 13 state-of-the-art vision-language models and observe a consistent performance drop as tasks shift from perceptual matching to logical contradiction detection. Most models perform well on isolated entity recognition but fail when multiple clues must be synthesized for conflict reasoning. Capability-level analysis further reveals uneven skill acquisition, especially in tasks requiring multi-step inference or rule-based validation.
Additional probing shows that conventional prompting strategies such as Chain-of-Thought and Set-of-Mark yield only marginal gains. By contrast, methods that interleave symbolic reasoning with grounded visual processing achieve more stable improvements. These results highlight a persistent bottleneck in multimodal reasoning and suggest new directions for building models capable of robust cross-modal verification.

Downloads

Next from AAAI 2026

Debiasing Machine Learning Predictions for Causal Inference Without Additional Ground Truth Data: “One Map, Many Trials” in Satellite-Driven Poverty Analysis

Similar lecture

Global-Local Confidence Fusion for Hallucination Detection in Mathematical Reasoning Task