United States

In the rapidly evolving landscape of large language models (LLMs) for medical applications, ensuring the reliability and accuracy of these models in clinical settings is paramount. Existing benchmarks often focus on fixed-format tasks like multiple-choice QA, which fail to capture the complexity of real-world clinical diagnostics. Moreover, traditional evaluation metrics and LLM-based evaluators struggle with misalignment, often providing oversimplified assessments that do not adequately reflect human judgment. To address these challenges, we introduce HDCEval, a **H**ierarchical **D**ivide-and-**C**onquer **Eval**uation framework tailored for fine-grained alignment in medical evaluation. HDCEval is built on a set of fine-grained medical evaluation guidelines developed in collaboration with professional doctors, encompassing Patient Question Relevance, Medical Knowledge Correctness, and Expression. The framework decomposes complex evaluation tasks into specialized subtasks, each evaluated by expert models trained through Attribute-Driven Token Optimization (ADTO) on a meticulously curated preference dataset. This hierarchical approach ensures that each aspect of the evaluation is handled with expert precision, leading to a significant improvement in alignment with human evaluators.

AAAI 2025

Hierarchical Divide-and-Conquer for Fine-Grained Alignment in LLM-Based Medical Evaluation

interpretability analysis and evaluation of nlp models

snlp

technical paper

We are pleased to announce the Thirty-Ninth AAAI Conference on Artificial Intelligence (AAAI-25), which will be held in Philadelphia, Pennsylvania at the Pennsylvania Convention Center from February 25 to March 4, 2025.

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-25 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.

### [Invited Speakers](https://aaai.org/conference/aaai/aaai-25/aaai-25-invited-speakers/)

Register [here](https://aaai.org/conference/aaai/aaai-25/registration/)

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-25 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.



Mixed Integer Linear Program (MILP) solvers are mostly built upon a Branch-and-Bound (B&B) algorithm, where the efficiency of traditional solvers heavily depends on hand-crafted heuristics for branching. The past few years have witnessed the increasing popularity of data-driven approaches to learn these heuristics automatically. However, the success of these methods heavily relies on the availability of high-quality demonstrations, which requires either the development of near-optimal heuristics or a time-consuming sampling process. This paper averts this challenge by proposing Suboptimal-Demonstration-Guided Reinforcement Learning (SORREL) for learning to branch. SORREL selectively learns from suboptimal demonstrations based on value estimation. It utilizes suboptimal demonstrations through both offline reinforcement learning on the demonstrations generated by suboptimal heuristics and self-imitation learning on past good experiences sampled by itself. Our experiments demonstrate its advanced performance in both branching quality and training efficiency over previous methods for various MILP problems.

SORREL: Suboptimal-Demonstration-Guided Reinforcement Learning for Learning to Branch

Beginning with Witkowski et al. (2021), recent work on forecasting competitions has addressed incentive problems with the common winner-take-all mechanism. Frongillo et al. (2021) propose a competition mechanism based on Multiplicative Weights, an online learning algorithm. They show that their mechanism selects an $\epsilon$-optimal forecaster with high probability using only $O(\log(n)/\epsilon^2)$ events. These works, together with all prior work on this problem thus far, assume that events are independent. We prove the first accuracy and approximate truthfulness guarantees for forecasting competitions with correlated events. To quantify correlation, we introduce a notion of block correlation, which allows each event to be strongly correlated with up to $b$ others and weakly correlated with the rest. We show that under distributions with this correlation, the Multiplicative Weights mechanism retains its $\epsilon$-optimal guarantee using $O(b^2 \log(n)/\epsilon^2)$ events. Our proof involves a novel concentration bound for correlated random variables which may be of broader interest.

Forecasting Competitions with Correlated Events

Assignment problems are a classic combinatorial optimization problem in which a group of agents must be assigned to a group of tasks such that maximum utility is achieved while satisfying assignment constraints. Given the utility of each agent completing each task, polynomial-time algorithms exist to solve a single assignment problem in its simplest form. However, in many modern-day applications such as satellite constellations, power grids, and mobile robot scheduling, assignment problems unfold over time, with the utility for a given assignment depending heavily on the state of the system. We apply multi-agent reinforcement learning to this problem, learning the value of assignments by bootstrapping from the known polynomial-time greedy solver and then learning from further experience. We then choose assignments using a distributed optimal assignment mechanism rather than by selecting them directly. We demonstrate that this algorithm is theoretically justified and avoids pitfalls experienced by other RL algorithms in this setting. Finally, we show that our algorithm significantly outperforms other methods in the literature, even while scaling to realistic scenarios with hundreds of agents and tasks.

Multi Agent Reinforcement Learning for Sequential Satellite Assignment Problems

Multi-modality pre-training paradigm that aligns protein sequences and biological descriptions has learned general protein representations and achieved promising performance in various downstream applications. However, these works were still unable to replicate the extraordinary success of language-supervised visual foundation models due to the ineffective usage of aligned protein-text paired data and the lack of an effective function-informed pre-training paradigm. To address these issues, this paper curates a large-scale protein-text paired dataset called ProtAnno with a property-driven sampling strategy, and introduces a novel function-informed protein pre-training paradigm. Specifically, the sampling strategy determines sampling probability based on the confidence and coverage of property annotations, balancing the data quality and data quantity in face of large-scale noisy data. Furthermore, motivated by significance of the protein specific functional mechanism, the proposed paradigm explicitly model protein static and dynamic functional segments by two segment-wise pre-training objectives, injecting fine-grained information in a function-informed manner. Leveraging all these innovations, we develop ProtCLIP, a multi-modality foundational model that comprehensively represents function-aware protein embeddings. On 22 different protein benchmarks within 5 types, including protein functionality classification, mutation effect prediction, cross-modality transformation, and protein-protein interaction prediction, our ProtCLIP consistently achieves the state-of-the-art performance, with remarkable improvements of 75% on average in five cross-modality transformation tasks, 59.9% in GO-CC and 39.7% in GO-BP protein function prediction. The experimental results demonstrate the extraordinary potential of ProtCLIP serving as the protein multi-modality foundation model.

ProtCLIP: Function-Informed Protein Multi-Modal Learning

Neural video compression has recently demonstrated significant potential to compete with conventional video codecs in terms of rate-quality performance. These learned video codecs are however associated with various issues related to decoding complexity (for autoencoder-based methods) and/or system delays (for implicit neural representation (INR) based models), which currently prevent them from being deployed in practical applications. In this paper, targeting a practical neural video codec, we propose a novel INR-based coding framework, PNVC, which innovatively combines autoencoder-based and overfitted solutions. Our approach benefits from several design innovations, including a new structural reparameterization-based architecture, hierarchical quality control, modulation-based entropy modeling, and scale-aware positional embedding. Supporting both low delay (LD) and random access (RA) configurations, PNVC outperforms existing INR-based codecs, achieving nearly 35\%+ BD-rate savings against HEVC HM 18.0 (LD) - almost 10\% more compared to one of the state-of-the-art INR-based codecs, HiNeRV and 5\% more over VTM 20.0 (LD), while maintaining 20+ FPS decoding speeds for 1080p content. This represents an important step forward for INR-based video coding, moving it towards practical deployment. The source code will be available for public evaluation.

PNVC: Towards Practical INR-based Video Compression

While machine learning (ML) models of today have the potential to be useful in many societal applications, they also harbor the potential for great harm, be it perpetuating biases or compromising privacy. To prevent these harms, many (evolving) regulatory guardrails have been put in place; for instance European Union's GDPR and Biden's Executive Order which demand explainability, privacy, fairness and so on from models deployed in societal applications. Yet, most technical solutions in the Trustworthy ML literature which claim to meet these regulatory requirements are brittle and often fail at the task in hand. To this end, my research aims to make the field of Trustworthy ML reliable using mainstay concepts of Measurement, Mitigation and Maintenance. With these concepts, I develop end-to-end solutions for trustworthy ML by (1) exploring the limitations of existing approaches and (2) providing principled novel solutions exploiting interconnections with cryptography.

The Mainstays of Trustworthy Machine Learning

While advances in machine learning and the expansion of massive datasets have significantly improved predictive accuracy, the translation of these predictions into actionable decisions—alongside a robust understanding of associated risks—remains underexplored. My research focuses on developing methodology and theory in data-driven decision-making and uncertainty quantification that effectively address core data challenges. This paper presents two connected pillars of my research: data-driven contextual optimization, uncertainty quantification and reduction.

Statistical Methodologies for Decision-Making and Uncertainty Reduction in Machine Learning

Deploying machine learning (ML) models in high-stakes domains such as healthcare and autonomous systems requires reliable uncertainty quantification (UQ) to ensure safe and accurate decision-making. 
Conformal prediction (CP) offers a robust, distribution-agnostic framework for UQ, providing valid prediction sets that guarantee a specified coverage probability. 
However, existing CP methods are often limited by assumptions that are violated in real-world scenarios, such as non-i.i.d. data, and by a lack of integration with modern machine learning workflows, particularly in large generative models. 
This research aims to address these limitations by advancing CP techniques to operate effectively in non-i.i.d. settings, improving predictive efficiency without sacrificing theoretical guarantees, and integrating CP directly into model training processes. 
These developments will enhance the practical applicability of CP for a wide range of ML tasks, enabling more reliable and interpretable models in high-stakes applications.

Reliable Uncertainty Quantification in Machine Learning via Conformal Prediction

Mechanistic interpretability is one of the most exciting and important research programs in current AI. My aim is to build some philosophical foundations for the program, along with setting out some concrete challenges and assessing progress to date. I will argue for the importance of propositional interpretability, which involves interpreting a system’s mechanisms and behavior in terms of propositional attitudes: attitudes (such as belief, desire, or subjective probability) to propositions (e.g. the proposition that it is hot outside). Propositional attitudes are the central way that we interpret and explain human beings and they are likely to be central in AI too. A central challenge is what I call thought logging: creating systems log all of the relevant propositional attitudes in an AI system over time. I will examine currently popular methods of interpretability (such as probing, sparse auto-encoders, and chain of thought methods) as well as philosophical methods of interpretation (including psychosemantics and representation theorems) to assess their strengths and weaknesses as methods of propositional interpretability.

Hierarchical Divide-and-Conquer for Fine-Grained Alignment in LLM-Based Medical Evaluation

Downloads

Next from AAAI 2025

SORREL: Suboptimal-Demonstration-Guided Reinforcement Learning for Learning to Branch

Stay up to date with the latest Underline news!

PRODUCT

COMPANY

RESOURCES

Hierarchical Divide-and-Conquer for Fine-Grained Alignment in LLM-Based Medical Evaluation

.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 2025

SORREL: Suboptimal-Demonstration-Guided Reinforcement Learning for Learning to Branch

Stay up to date with the latest Underline news!

PRODUCT

COMPANY

RESOURCES

Downloads