Singapore

Reinforcement Learning with Verifiable Rewards (RLVR) significantly enhances the reasoning capability of Large Language Models (LLMs). Current RLVR approaches typically conduct training across all generated tokens, but neglect to explore which tokens (e.g., prefix tokens) actually contribute to reasoning. This uniform training strategy spends substantial effort on optimizing low-return tokens, which in turn impedes the potential improvement from high-return tokens and reduces overall training effectiveness. To address this issue, we propose a novel RLVR approach called Progressive Prefix-token Policy Optimization (PPPO), which highlights the significance of the prefix segment of generated outputs. Specifically, inspired by the well-established human thinking theory of Path Dependence, where early-stage thoughts substantially constrain subsequent thinking trajectory, we identify an analogous phenomenon in LLM reasoning termed Beginning Lock-in Effect (BLE). PPPO leverages this finding by focusing its optimization objective on the prefix reasoning process of LLMs. This targeted optimization strategy can positively influence subsequent reasoning processes, and ultimately improve final results. To improve the learning effectiveness of LLMs on how to start reasoning with high quality, PPPO introduces two training strategies: (a) Progressive Prefix Retention, which shapes a progressive learning process by increasing the proportion of retained prefix tokens during training; (b) Continuation Accumulated Reward, which mitigates reward bias by sampling multiple continuations for one prefix token sequence, and accumulating their scores as the reward signal. Extensive experimental results on various reasoning tasks (e.g., math, physics, chemistry, and biology) demonstrate that our proposed PPPO outperforms representative RLVR methods, with the accuracy improvements of 18.02% on only 26.17% training tokens.

AAAI 2026

Well Begun, Half Done: Reinforcement Learning with Prefix Optimization for LLM Reasoning

simulating human behavior

(large) language models

reinforcement learning

### Welcome!
The 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024) will take place in Bangkok, Thailand from August 11th to 16th, 2024. Our Virtual Poster Sessions will take place online Thursday, August 22, 2024.

You are required to register for this event. **Please register [here](https://2024.aclweb.org/registration). **

If you have already registered, please check your inbox for an email from Underline granting you access to ACL 2024 content.

Please register!

ACL 2024

The 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024) will take place in Bangkok, Thailand from August 11th to 16th, 2024. More information will be announced soon.

poster

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.

Malware detection in real-world settings must deal with evolving threats, limited labeling budgets, and uncertain predictions. Traditional classifiers, without additional mechanisms, struggle to maintain performance under concept drift in malware domains, as their supervised learning formulation cannot optimize when to defer decisions to manual labeling and adaptation. Modern malware detection pipelines combine classifiers with monthly active learning (AL) and rejection mechanisms to mitigate the impact of concept drift. In this work, we develop a novel formulation of malware detection as a one-step Markov Decision Process and train a deep reinforcement learning (DRL) agent, simultaneously optimizing sample classification performance and rejecting high-risk samples for manual labeling. We evaluated the joint detection and drift mitigation policy learned by the DRL-based Malware Detection (DRMD) agent through time-aware evaluations on Android malware datasets subject to realistic drift requiring multi-year performance stability. The policies learned under these conditions achieve a higher Area Under Time (AUT) performance compared to standard classification approaches used in the domain, showing improved resilience to concept drift. Specifically, the DRMD agent achieved a $5.18\pm5.44$, $14.49\pm12.86$, and $10.06\pm10.81$ average AUT performance improvement for the classification only, classification with rejection, and classification with rejection and AL settings, respectively. Our results demonstrate for the first time that DRL can facilitate effective malware detection and improved resiliency to concept drift in the dynamic environment of the Android malware domain.

DRMD: Deep Reinforcement Learning for Malware Detection Under Concept Drift

Heterogeneous agent reinforcement learning (HARL) enables intelligent agents to execute complex cooperative tasks by adopting agent-specific policies. Existing HARL methods often use distinct networks per agent to ensure monotonic improvement, leading to substantial computational overhead and limited scalability in resource-constrained environments. To overcome this limitation, we propose SDE-HARL to scale HARL to a large number of agents while maintaining effective inter-agent coordination. Specifically, SDE-HARL decomposes the policy network of each agent into a lightweight local network and a global network. As such, our proposed method enables efficient local computing while allowing agent-specific properties. Moreover, to achieve efficient adaptation, agents with similar roles are grouped via a role-aware mechanism and share partial parameters in their global networks, while an identity-aware mechanism is introduced to promote behavioral diversity among agents within the same group. In certain scenarios across Google Research Football and StarCraft II, SDE-HARL reaches about 90% win rate while halving inference time compared to standard architectures.

SDE-HARL: Scalable Distributed Policy Execution for Heterogeneous-Agent Reinforcement Learning

Artificial Intelligence systems are increasingly deployed in settings where ensuring robustness, fairness, or domain-specific properties is essential for regulation compliance and alignment with human values. However, especially on Neural Networks, property enforcement is very challenging, and existing methods are limited to specific constraints or local properties (defined around datapoints), or fail to provide full guarantees. We tackle these limitations by extending SMiLE, a recently proposed enforcement framework for NNs, to support global relational properties (defined over the entire input space). The proposed approach scales well with model complexity, accommodates general properties and backbones, and provides full satisfaction guarantees. We evaluate SMiLE on monotonicity, global robustness, and individual fairness, on synthetic and real data, for regression and classification tasks. Our approach is competitive with property-specific baselines in terms of accuracy and runtime, and strictly superior in terms of generality and level of guarantees. Overall, our results emphasize the potential of the SMiLE framework as a platform for future research and applications.

SMiLE: Provably Enforcing Global Relational Properties in Neural Networks

Current Video Large Language Models (VideoLLMs) suffer from quadratic computational complexity and key-value cache scaling, due to their reliance on processing excessive redundant visual tokens. To address this problem, we propose SharpV, a minimalist and efficient method for adaptive pruning of visual tokens and KV cache. Different from most uniform compression approaches, SharpV dynamically adjusts pruning ratios based on spatial-temporal information. Remarkably, this adaptive mechanism occasionally achieves performance gains over dense models, offering a novel paradigm for adaptive pruning. During the KV cache pruning stage, based on observations of visual information degradation, SharpV prunes degraded visual features via a self-calibration manner, guided by similarity to original visual features. In this way, SharpV achieves hierarchical cache pruning from the perspective of information bottleneck, offering a new insight into VideoLLMs' information flow. Experiments on multiple public benchmarks demonstrate the superiority of SharpV. Moreover, to the best of our knowledge, SharpV is notably the first two-stage pruning framework that operates without requiring access to exposed attention scores, ensuring full compatibility with hardware acceleration techniques like Flash Attention.

Sharp Eyes and Memory for VideoLLMs: Information-Aware Visual Token Pruning for Efficient and Reliable VideoLLM Reasoning

The transfer of knowledge from large-scale pre-trained models to diverse downstream tasks has achieved remarkable success. Beyond the traditional full fine-tuning paradigm, Parameter-Efficient Fine-Tuning (PEFT) has emerged as a more efficient model adaptation approach. However, applying existing PEFT methods to adapt dense vision models, particularly in multi-task settings, remains inadequately explored due to their low efficiency, limited task scalability, and neglect of cross-task fine-tuning interactions. To address these challenges, we propose the Task Dynamic-Synergistic Skill Adaptation, termed TDSS, an efficient and scalable multi-task model adaptation framework for dense visual predictions. TDSS comprises two key components: Task-Dynamic Skill Adapters (TDSA) and Task-Synergistic Adaptation Interaction (TSAI). Specifically, TDSA are inserted in parallel into pre-trained vision models to extract task-specific adapted features through the construction of skill representation experts and task dynamic gating. TSAI is developed to enhance cross-task adaptation interaction by bridging global generic and task-specific adapted features. Extensive experiments on multi-task dense visual predictions demonstrate that TDSS surpasses existing state-of-the-art parameter-efficient fine-tuning methods, while exhibiting remarkable efficiency and scalability in parameters and computational complexity.

TDSS: Task Dynamic-Synergistic Skill Adaptation for Boosting Efficient and Scalable Multi-Task Learning in Dense Visual Prediction

Recent advances in long-context language models (LCLMs), designed to handle extremely long contexts, primarily focus on utilizing external contextual information, often leaving the influence of language models' parametric knowledge underexplored. In this work, we first investigate how this parametric knowledge affects content generation and demonstrate that its impact becomes increasingly pronounced as context length extends. Furthermore, we show that the model’s ability to utilize parametric knowledge, which we call parametric recall ability, does not improve simultaneously with its ability to leverage contextual knowledge through extrinsic retrieval ability. Moreover, better extrinsic retrieval ability can interfere with the model’s parametric recall ability, limiting its full potential. To bridge this gap, we design a simple yet effective Hybrid Needle-in-a-Haystack test that evaluates models based on their capabilities across both abilities, rather than solely emphasizing extrinsic retrieval ability. Our experimental results reveal that Qwen-2.5 models significantly outperform Llama-3.1 models, demonstrating superior potential to combine various abilities. Moreover, even the more powerful Llama-3.1-70B-Instruct model fails to exhibit better performance, highlighting the importance of evaluating models from a dual-ability perspective.

Harnessing the Unseen: The Hidden Influence of Intrinsic Knowledge in Long-Context Language Models

Recent advances in Multi-modal Large Language Models (MLLMs) have showcased remarkable capabilities in vision-language understanding. However, enabling robust video spatial reasoning—the ability to comprehend object locations, orientations, and inter-object relationships in dynamic 3D scenes—remains a key unsolved challenge. Existing approaches primarily rely on spatially grounded supervised fine-tuning or reinforcement learning, yet we observe that such models often exhibit query-locked reasoning, focusing narrowly on objects explicitly mentioned in the prompt while ignoring critical contextual cues. To address this limitation, we propose Object-Centric 3D Rollout (OCR), a novel strategy that introduces structured perturbations to the 3D geometry of selected objects during training. By degrading object-specific visual cues and projecting the altered geometry into 2D space, OCR compels the model to reason holistically across the entire scene. We further design a rollout-based training pipeline that jointly leverages vanilla and region-noisy videos to optimize spatial reasoning trajectories. Experiments demonstrate state-of-the-art performance: our 3B-parameter model achieves 47.5\% accuracy on VSI-Bench, outperforming several 7B baselines. Ablations confirm OCR’s superiority over prior rollout strategies (e.g., T-GRPO, NoisyRollout).

Video Spatial Reasoning with Object-Centric 3D Rollout

Neural operators have shown great potential in solving a family of Partial Differential Equations (PDEs) by modeling the mappings between input and output functions. Fourier Neural Operator (FNO) implements global convolutions via parameterizing the integral operators in Fourier space. However, it often results in over-smoothing solutions and fails to capture local details and high-frequency components. To address these limitations, we investigate incorporating the spatial-frequency localization property of Wavelet transforms into the Transformer architecture. We propose a novel Wavelet Attention (WA) module with linear computational complexity to efficiently learn locality-aware features. Building upon WA, we further develop the Spectral Attention Operator Transformer (SAOT), a hybrid spectral Transformer framework that integrates WA’s localized focus with the global receptive field of Fourier-based Attention (FA) through a gated fusion block. Experimental results demonstrate that WA significantly mitigates the limitations of FA and outperforms existing Wavelet-based neural operators by a large margin. By integrating the locality-aware and global spectral representations, SAOT achieves state-of-the-art performance on six operator learning benchmarks and exhibits strong discretization-invariant ability.

SAOT: An Enhanced Locality-Aware Spectral Transformer for Solving PDEs

The rapid advancement of diffusion-based image editing has enabled highly controllable visual content generation but has also raised serious concerns about the misuse of generative models for producing Not-Safe-for-Work (NSFW) content. Existing protection strategies inject adversarial perturbations to disrupt editing. However, these methods are untargeted, often degrading benign edits while failing to eliminate harmful outputs. In this work, we propose TarPro, a targeted protection framework that blocks malicious edits while preserving benign editing functionality. TarPro introduces Dual-Intent Optimization (DIO), a semantic alignment objective that suppresses malicious prompt effects while retaining desirable, benign edits, by leveraging prompt compositionality rather than requiring manually annotated preferences. To ensure robustness and generalization, we replace pixel-level optimization with a generator-based perturbation learning strategy that learns to produce structured, imperceptible perturbations in parameter space. Experiments on multiple diffusion backbones show that TarPro significantly blocks NSFW content while maintaining high-quality benign edits, outperforming strong baselines through both qualitative and quantitative evaluations.

TarPro: Targeted Protection Against Malicious Image Editing

Structure optimization, which yields the relaxed structure (minimum‑energy state), is essential for reliable materials property calculations, yet traditional ab initio approaches such as density‑functional theory (DFT) are computationally intensive. Machine learning (ML) has emerged to alleviate this bottleneck but suffers from two major limitations: (i) existing models operate mainly on atoms, leaving lattice vectors implicit despite their critical role in structural optimization; and (ii) they often rely on multi-stage, non-end-to-end workflows that are prone to error accumulation. Here, we present E$^{3}$Relax—an end-to-end SE(3)‑equivariant graph neural network that maps an unrelaxed crystal directly to its relaxed structure. E$^{3}$Relax promotes both atoms and lattice vectors to graph nodes endowed with dual scalar–vector features, enabling unified, symmetry‑preserving modeling of atomic displacements and lattice deformations. A layer‑wise supervision strategy forces every network depth to make a physically meaningful refinement, mimicking the incremental convergence of DFT while preserving a fully end‑to‑end pipeline. We evaluate E$^{3}$Relax on four benchmark datasets and demonstrate that it achieves remarkable accuracy and efficiency. Through DFT validations, we show that the structures predicted by E$^{3}$Relax are energetically favorable, making them suitable as high-quality initial configurations to accelerate DFT calculations.

Downloads

Next from AAAI 2026

DRMD: Deep Reinforcement Learning for Malware Detection Under Concept Drift

Similar lecture

Selective Prefix Tuning for Pre-trained Language Models

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from AAAI 2026

DRMD: Deep Reinforcement Learning for Malware Detection Under Concept Drift

Similar lecture

Selective Prefix Tuning for Pre-trained Language Models

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES

Downloads