Singapore

Large language models now draft news, legal analyses, and software code with human-level fluency. At the same time, regulations such as the EU AI Act mandate that each synthetic passage carry an imperceptible, machine-verifiable mark for provenance. Conventional logit-based watermarks satisfy this requirement by selecting a pseudorandom green vocabulary at every decoding step and boosting its logits, yet the random split can exclude the highest-probability token and thus erode fluency. WaterMod mitigates this limitation through a probability-aware modular rule. The vocabulary is first sorted in descending model probability; the resulting ranks are then partitioned by the residue $\text{rank}\bmod k$, which distributes adjacent—and therefore semantically similar—tokens across different classes.
A fixed bias of small magnitude is applied to one selected class. In the zero-bit setting ($k=2$), an entropy-adaptive gate selects either the even or the odd parity as the green list. Because the top two ranks fall into different parities, this choice embeds a detectable signal 
while guaranteeing that at least one high-probability token remains available for sampling. In the multi-bit regime ($k&gt;2$), the current payload digit $d$ selects the color class whose ranks satisfy $\text{rank} \bmod k = d$. Biasing the logits of that class embeds exactly 
one base-$k$ digit—equivalently $\log_{2}k$ bits—per decoding step, thereby enabling fine-grained provenance tracing. The same modular arithmetic therefore supports both binary attribution and rich payloads. Experimental results demonstrate that WaterMod consistently attains strong watermark detection performance while maintaining generation quality in both zero-bit and multi-bit settings. This robustness holds across a range of tasks, including natural language generation, mathematical reasoning, and code synthesis. Our code and data are available at \url{https://github.com/Shinwoo-Park/WaterMod}.

AAAI 2026

WaterMod: Modular Token-Rank Partitioning for Probability-Balanced LLM Watermarking

text watermarking

It is my great pleasure and honor to welcome you to the 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026), held in San Diego, California, USA, July 2–7, 2026. ACL 2026 continues our fields tradition of excellence in scholarship, innovation, and inclusivity, and I am deeply grateful to the many volunteers who have worked tirelessly to bring this event to life.

It is exciting to see the many developments in the field reflected at our conference, such as the advancement of large language models (born out of our community) into many application directions. This is not without challenges, partly due to unsolved problems, and partly due to misuse of the technology. All this is reflected well at ACL 2026. The conference also reflects that the field is not only about advancing technology, but also the science of fundamentally understanding language and how it is use by many different speaker communities. 
[Full Message from General Chair Philipp Koehn](https://drive.google.com/file/d/1QiJFOmMjvDgsDo3i1XzlQnImxe3RU_Cy/view?usp=sharing) 
<html><button style="display: inline-flex; align-items: center; justify-content: center; white-space: nowrap; border-radius: 9999px; font-weight: bold; background: #7c3aed; color: white; font-family: 'Space Grotesk', sans-serif; height: 40px; font-size: 16px; padding: 0 20px; border: none; cursor: pointer" onclick="window.open('https://drive.google.com/file/d/1QiJFOmMjvDgsDo3i1XzlQnImxe3RU_Cy/view?usp=sharing')">Conference Handbook</button></html>

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 ACL 2026". Please be sure to check your spam email folder if you do not see an email confirmation right away.

To access this event page, you are required to register.
Please complete your registration to continue.

We recommend reading [**the registration information**](https://2026.aclweb.org/registration/) first.

**Online Registration Form**: https://acl.swoogo.com/acl2026

Registration Required

ACL 2026

The 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026) will take place in San Diego, California, from July 2 - 7, 2026.

Large language models now draft news, legal analyses, and software code with human-level fluency. At the same time, regulations such as the EU AI Act mandate that each synthetic passage carry an imperceptible, machine-verifiable mark for provenance. Conventional logit-based watermarks satisfy this requirement by selecting a pseudorandom green vocabulary at every decoding step and boosting its logits, yet the random split can exclude the highest-probability token and thus erode fluency. WaterMod mitigates this limitation through a probability-aware modular rule. The vocabulary is first sorted in descending model probability; the resulting ranks are then partitioned by the residue $\text{rank}\bmod k$, which distributes adjacent—and therefore semantically similar—tokens across different classes.
A fixed bias of small magnitude is applied to one selected class. In the zero-bit setting ($k=2$), an entropy-adaptive gate selects either the even or the odd parity as the green list. Because the top two ranks fall into different parities, this choice embeds a detectable signal 
while guaranteeing that at least one high-probability token remains available for sampling. In the multi-bit regime ($k>2$), the current payload digit $d$ selects the color class whose ranks satisfy $\text{rank} \bmod k = d$. Biasing the logits of that class embeds exactly 
one base-$k$ digit—equivalently $\log_{2}k$ bits—per decoding step, thereby enabling fine-grained provenance tracing. The same modular arithmetic therefore supports both binary attribution and rich payloads. Experimental results demonstrate that WaterMod consistently attains strong watermark detection performance while maintaining generation quality in both zero-bit and multi-bit settings. This robustness holds across a range of tasks, including natural language generation, mathematical reasoning, and code synthesis. Our code and data are available at \url{https://github.com/Shinwoo-Park/WaterMod}.

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 

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.

Federated learning (FL) allows for collaborative model training while preserving data privacy, but its distributed nature makes it vulnerable to poisoning attacks. Existing defense methods typically rely on using gradients from multiple clients to define a trusted region, selecting only the trustworthy update (good gradients) within this region for aggregation. Mainstream defense boundaries are categorized as hard
boundaries, soft boundaries, and semi-soft boundaries. However, we argue that even good gradients within these boundaries can still be exploited by attackers to poison the model. To tackle this challenge, we introduce a boundary-adaptive attack method that leverages the directional properties of optimization techniques to derive baseline poisoned gradients. Through iterative perturbation, it generates seemingly innocent gradients that subtly deviate from the global model. Our extensive study on 3 benchmark datasets and 13 mainstream defensive mechanisms confirms that the proposed attack raises a significantly severe threat to the integrity and security of federated learning practices, regardless of the flourishing of robust Federated Learning methods.

Good Gradients Poison Your Model: Evading Defenses in Federated Learning via Boundary-adaptive Perturbation

Session-based recommendation (SBR) aims to predict anonymous users' next interaction based on their interaction sessions. In practical recommendation scenario, low-exposure items constitute the majority of interactions, creating a long-tail distribution that severely compromises recommendation diversity. Existing approaches attempt to address this issue by promoting tail items but incur accuracy degradation, exhibiting a "see-saw" effect between long-tail and accuracy performance. We attribute such conflict to session-irrelevant noise within the tail item set, which existing long-tail approaches fail to identify and constrain effectively. To resolve our fundamental conflict, we propose HID (Hybrid Intent-based Dual Constraint Framework), a plug-and-play framework that transforms the conventional "see-saw" into a "win-win" relationship through introducing the hybrid intent-based dual constraints. Two key innovations are incorporated in this framework: (i) Hybrid Intent Learning, where we reformulate the intent extraction strategies by employing attribute-aware spectral clustering to reconstruct the item-to-intent mapping. Furthermore, discrimination of session-irrelevant noise is achieved through the assignment of both target and noise intents to each sessions. (ii) Intent Constraint Loss, where we propose two novel constraint paradigms regarding the diversity and accuracy to regulate the representation learning process, and unify the two optimization objectives into a unique loss. Extensive experiments across multiple SBR models and datasets demonstrate that HID can enhance both long-tail performance and recommendation accuracy, establishing new state-of-the-art performance in long-tail recommender systems.

Bid Farewell to Seesaw: Towards Accurate Long-Tail Session-Based Recommendation via Dual Constraints of Hybrid Intents

Real world systems evolve in continuous-time according to their underlying causal relationships, yet their dynamics are often unknown. Existing approaches to learning such dynamics typically either discretize time ---leading to poor performance on irregularly sampled data--- or ignore the underlying causality. We propose CADYT, a novel method for causal discovery on dynamical systems addressing both these challenges. In contrast to state-of-the-art causal discovery methods that model the problem using discrete-time Dynamic Bayesian networks, our formulation is grounded in Difference-based causal models, which allow milder assumptions for modeling the continuous nature of the system. CADYT leverages exact Gaussian Process inference for modeling the continuous-time dynamics which is more aligned with the underlying dynamical process. We propose a practical instantiation that identifies the causal structure via a greedy search guided by the Algorithmic Markov Condition and Minimum Description Length principle. Our experiments show that CADYT outperforms state-of-the-art methods on both regularly and irregularly-sampled data, discovering causal networks closer to the true underlying dynamics.

Causal Structure Learning for Dynamical Systems with Theoretical Score Analysis

Streamlining constraints (or streamliners, for short) narrow the search space, enhancing the speed and feasibility of solving complex constraint satisfaction problems. Traditionally, streamliners were crafted manually or generated through systematically combined atomic constraints with high-effort offline testing. Our approach utilizes the generative capabilities of Large Language Models (LLMs) to propose effective streamliners for problems specified in the MiniZinc constraint programming language and integrates feedback to the LLM with quick empirical tests for validation. Evaluated across seven diverse constraint satisfaction problems, our method achieves substantial runtime reductions. We compare the results to obfuscated and disguised variants of the problem to see whether the results depend on LLM memorization. We also analyze whether longer offline runs improve the quality of streamliners and whether the LLM can propose good combinations of streamliners.

Generating Streamlining Constraints with Large Language Models

3D medical image fusion (MIF) and segmentation (MIS) are critical and inherently synergistic tasks in medical image analysis. However, fundamentally integrating them remains highly challenging, since effective collaborative paradigms are still scarce and their optimization objectives fundamentally diverge. Moreover, existing continual learning techniques are unable to achieve truly advanced performance for both tasks using a shared weight. To address these challenges, we propose M²-CoFS, a unified model capable of jointly handling both tasks. Our core contribution is a “network-guided network learning” paradigm designed to break the task boundaries. We model the weight spaces of MIF and MIS as high-dimensional manifolds and innovatively use a lightweight neural network to implicitly construct a shared manifold. Interestingly, this network yields a unified weight for both tasks. To ensure the shared manifold retains the intrinsic geometry of both original manifolds, we embed manifold distances into the loss function of this network as a constraint. Additionally, we design a tailored three-stage training paradigm for our core contribution mentioned above. Stage I focuses on independent task optimization for high-quality weights; Stage II aims to reduce parameter-space distance between tasks via our cross-task weight adaptation strategy; Our core innovation serves as stage III. Experimental results show that M²-CoFS consistently outperforms state-of-the-art comparison models on both MlF and MIS.

Breaking Task Boundaries: A Unified Model for 3D Medical Image Fusion and Segmentation Guided by Manifold Perspective

A fundamental use of knowledge bases (KBs) is query answering, i.e.,
retrieving the information entailed by the KB in response to a user
query. When both the KB and the query are specified as logical
formulae, the standard form of answer provided to users is the set
of all certain answers (CAs): tuples of constants that satisfy the
formula defining the query in every model of the logical theory
defining the KB.

Despite their wide adoption, CAs are known to be just a lossy
representation of the information that a KB and a query provide.
While several alternative answer languages have been proposed in the
literature, no general consensus has emerged on the most suitable
approach to query answering over ontological KBs, as each language
come with its own limitations.

To address some of these issues, we introduce Regularly Recurrent
Answers (RRAs), a novel answer language for queries over
ontological KBs based on regular expressions. RRAs support the
representation of infinite sets of tuples of constants via a simple
(and arguably well understood) generation mechanism. We show that
RRAs can capture a fundamental fragment of the certain information
entailed by union of conjunctive queries and DL-Lite KBs, making
them a strong candidate for informative query answering
settings. Our contribution include the formal definition of RRAs, a
proof of their informativeness, and a study of the computational
complexity of query answering problem using RRAs.

Expressive Recursive Answers for Ontological Knowledge Bases

Hard negatives are essential for training effective retrieval models. Hard-negative mining typically relies on ranking documents using cross-encoders or static embedding models based on similarity metrics such as cosine distance. Hard negative mining becomes challenging for biomedical and scientific domains due to the difficulty in distinguishing between source and hard negative documents. However, referenced documents naturally share contextual relevance with the source document but are not duplicates, making them well-suited as hard negatives. In this work, we propose BiCA: Biomedical Dense Retrieval with Citation-Aware Hard Negatives, an approach for hard-negative mining by utilizing citation links in 20,000 PubMed articles for improving a domain-specific small dense retriever. We fine-tune the GTE_small and GTE_Base models using these citation-informed negatives and observe consistent improvements in zero-shot dense retrieval using nDCG@10 for both in-domain and out-of-domain tasks on BEIR and outperform baselines on long-tailed topics in LoTTE using Success@5. Our findings highlight the potential of leveraging document link structure to generate highly informative negatives, enabling state-of-the-art performance with minimal fine-tuning and demonstrating a path towards highly data-efficient domain adaptation.

BiCA: Effective Biomedical Dense Retrieval with Citation-Aware Hard Negatives

In-context learning (ICL) has proven to be adept at adapting large language models (LLMs) to downstream tasks without parameter updates, based on a few demonstration examples.
Prior work has found that the ICL performance is susceptible to the selection of examples in prompt and made efforts to stabilize it.
However, existing example selection studies ignore the ethical risks behind the examples selected, such as gender and race bias.
In this work, we conduct extensive experiments and discover that (1) example selection with high accuracy does not mean low bias; 
(2) example selection for ICL may amplify the biases of LLMs;
(3) example selection exacerbates some spurious correlations of LLMs.
To mitigate the bias amplification caused by selected examples, we propose Remind with Bias-aware Embedding (*ReBE*), which disrupts spurious correlations through contrastive learning and learns bias-aware embedding for LLMs via prompt tuning.
Finally, we demonstrate that ReBE effectively mitigates biases of LLMs without significantly compromising accuracy and is highly compatible with existing example selection methods.

The Silent Amplifier: In-Context Examples Fuel Bias in Large Language Models

Recent vision-language-action (VLA) models built on pretrained vision-language models (VLMs) have demonstrated strong performance in robotic manipulation. However, these models remain constrained by the single-frame image paradigm and fail to fully leverage the temporal information offered by multi-frame histories, as directly feeding multiple frames into VLM backbones incurs substantial computational overhead and inference latency. We propose CronusVLA, a unified framework that extends single-frame VLA models to the multi-frame paradigm. CronusVLA follows a two-stage process: (1) Single-frame pretraining on large-scale embodied datasets with autoregressive prediction of action tokens, establishing an effective embodied vision-language foundation; (2) Multi-frame post-training, which adapts the vision-language backbone’s prediction from discrete tokens to learnable features, and aggregates historical information via feature chunking. CronusVLA effectively addresses the existing challenges of multi-frame modeling while enhancing performance and observational robustness. To evaluate the robustness under temporal and spatial disturbances, we introduce SimplerEnv-OR, a novel benchmark featuring 24 types of observational disturbances and 120 severity levels. Experiments across three embodiments in simulated and real-world environments demonstrate that CronusVLA achieves state-of-the-art performance and superior robustness, with a 70.9\% success rate on SimplerEnv, a 17.2\% improvement over OpenVLA on LIBERO, and the highest robustness score on SimplerEnv-OR. These results highlight the potential of efficient multi-frame adaptation in VLA models for more powerful and robust real-world deployment.

Towards Efficient and Robust Manipulation via Multi-Frame Vision-Language-Action Modeling

A few pixel attack misleads a classifier by modifying a few pixels of an image. Its perturbation space is an $\ell_0$-ball, which is not convex, unlike $\ell_p$-balls for $p\geq1$. However, existing local robustness verifiers typically scale by relying on linear bound propagation, which captures convex perturbation spaces. We show that the convex hull of an $\ell_0$-ball is the intersection of its bounding box and an asymmetrically scaled $\ell_1$-like polytope. For single channel inputs, the volumes of the convex hull and this polytope are nearly equal as the input dimension increases. We then show a linear bound propagation that precisely computes bounds over the convex hull and is significantly tighter than bound propagations over the bounding box or our $\ell_1$-like polytope. This bound propagation scales the state-of-the-art $\ell_0$ verifier on robustness benchmarks by 1.24x-7.07x, with geometric mean of 3.16.

Downloads

Next from AAAI 2026

Good Gradients Poison Your Model: Evading Defenses in Federated Learning via Boundary-adaptive Perturbation

Similar lecture

SSG: Logit-Balanced Vocabulary Partitioning for LLM Watermarking