Neural Theorem Proving for Verification Conditions: A Real-World Benchmark
arXiv:2601.18944v1 Announce Type: new
Abstract: Theorem proving is fundamental to program verification, where the automated proof of Verification Conditions (VCs) remains a primary bottleneck. Real-world program verification frequently encounters hard VCs that existing Automated Theorem Provers (AT...
arXiv:2601.16984v1 Announce Type: new
Abstract: The 3rd Generation Partnership Project (3GPP) produces complex technical specifications essential to global telecommunications, yet their hierarchical structure, dense formatting, and multi-modal content make them difficult to process. While Large Lan...
Sparsity-Aware Low-Rank Representation for Efficient Fine-Tuning of Large Language Models
arXiv:2601.16991v1 Announce Type: new
Abstract: Adapting large pre-trained language models to downstream tasks often entails fine-tuning millions of parameters or deploying costly dense weight updates, which hinders their use in resource-constrained environments. Low-rank Adaptation (LoRA) reduces ...
MathMixup: Boosting LLM Mathematical Reasoning with Difficulty-Controllable Data Synthesis and Curriculum Learning
arXiv:2601.17006v1 Announce Type: new
Abstract: In mathematical reasoning tasks, the advancement of Large Language Models (LLMs) relies heavily on high-quality training data with clearly defined and well-graded difficulty levels. However, existing data synthesis methods often suffer from limited di...
Interpreting Agentic Systems: Beyond Model Explanations to System-Level Accountability
arXiv:2601.17168v1 Announce Type: new
Abstract: Agentic systems have transformed how Large Language Models (LLMs) can be leveraged to create autonomous systems with goal-directed behaviors, consisting of multi-step planning and the ability to interact with different environments. These systems diff...
High-Fidelity Longitudinal Patient Simulation Using Real-World Data
arXiv:2601.17310v1 Announce Type: new
Abstract: Simulation is a powerful tool for exploring uncertainty. Its potential in clinical medicine is transformative and includes personalized treatment planning and virtual clinical trials. However, simulating patient trajectories is challenging because of ...
arXiv:2601.17311v1 Announce Type: new
Abstract: Multi-agent systems can improve reliability, yet under a fixed inference budget they often help, saturate, or even collapse. We develop a minimal and calibratable theory that predicts these regimes from three binding constraints of modern agent stacks...
Principled Coarse-Grained Acceptance for Speculative Decoding in Speech
Speculative decoding accelerates autoregressive speech generation by letting a fast draft model propose tokens that a larger target model verifies. However, for speech LLMs that generate acoustic tokens, exact token matching is overly restrictive: many discrete tokens are acoustically or semanticall...
SelfReflect: Can LLMs Communicate Their Internal Answer Distribution?
The common approach to communicate a large language model’s (LLM) uncertainty is to add a percentage number or a hedging word to its response. But is this all we can do? Instead of generating a single answer and then hedging it, an LLM that is fully transparent to the user needs to be able to reflec...
Learning to Reason as Action Abstractions with Scalable Mid-Training RL
Large language models excel with reinforcement learning (RL), but fully unlocking this potential requires a mid-training stage. An effective mid-training phase should identify a compact set of useful actions and enable fast selection among them through online RL. We formalize this intuition by prese...
VLSU: Mapping the Limits of Joint Multimodal Understanding for AI Safety
Safety evaluation of multimodal foundation models often treats vision and language inputs separately, missing risks from joint interpretation where benign content becomes harmful in combination. Existing approaches also fail to distinguish clearly unsafe content from borderline cases, leading to pro...
Analyzing Neural Network Information Flow Using Differential Geometry
arXiv:2601.16366v1 Announce Type: new
Abstract: This paper provides a fresh view of the neural network (NN) data flow problem, i.e., identifying the NN connections that are most important for the performance of the full model, through the lens of graph theory. Understanding the NN data flow provide...
When Agents Fail to Act: A Diagnostic Framework for Tool Invocation Reliability in Multi-Agent LLM Systems
arXiv:2601.16280v1 Announce Type: new
Abstract: Multi-agent systems powered by large language models (LLMs) are transforming enterprise automation, yet systematic evaluation methodologies for assessing tool-use reliability remain underdeveloped. We introduce a comprehensive diagnostic framework tha...
SemanticALLI: Caching Reasoning, Not Just Responses, in Agentic Systems
arXiv:2601.16286v1 Announce Type: new
Abstract: Agentic AI pipelines suffer from a hidden inefficiency: they frequently reconstruct identical intermediate logic, such as metric normalization or chart scaffolding, even when the user's natural language phrasing is entirely novel. Conventional boundar...
DSGym: A Holistic Framework for Evaluating and Training Data Science Agents
arXiv:2601.16344v1 Announce Type: new
Abstract: Data science agents promise to accelerate discovery and insight-generation by turning data into executable analyses and findings. Yet existing data science benchmarks fall short due to fragmented evaluation interfaces that make cross-benchmark compari...
Doc2AHP: Inferring Structured Multi-Criteria Decision Models via Semantic Trees with LLMs
arXiv:2601.16479v1 Announce Type: new
Abstract: While Large Language Models (LLMs) demonstrate remarkable proficiency in semantic understanding, they often struggle to ensure structural consistency and reasoning reliability in complex decision-making tasks that demand rigorous logic. Although class...
SycoEval-EM: Sycophancy Evaluation of Large Language Models in Simulated Clinical Encounters for Emergency Care
arXiv:2601.16529v1 Announce Type: new
Abstract: Large language models (LLMs) show promise in clinical decision support yet risk acquiescing to patient pressure for inappropriate care. We introduce SycoEval-EM, a multi-agent simulation framework evaluating LLM robustness through adversarial patient ...
AI Can’t Do Physics Well – And That’s a Roadblock to Autonomy
QuantiPhy is a new benchmark and training framework that evaluates whether AI can numerically reason about physical properties in video images. QuantiPhy reveals that today’s models struggle with basic estimates of size, speed, and distance but offers a way forward.
Researchers tested AI against 100,000 humans on creativity
A massive new study comparing more than 100,000 people with today’s most advanced AI systems delivers a surprising result: generative AI can now beat the average human on certain creativity tests. Models like GPT-4 showed strong performance on tasks designed to measure original thinking and idea gen...
Empowering LLMs for Structure-Based Drug Design via Exploration-Augmented Latent Inference
arXiv:2601.15333v1 Announce Type: new
Abstract: Large Language Models (LLMs) possess strong representation and reasoning capabilities, but their application to structure-based drug design (SBDD) is limited by insufficient understanding of protein structures and unpredictable molecular generation. T...
arXiv:2601.15337v1 Announce Type: new
Abstract: Users should not be systemically disadvantaged by the language they use for interacting with LLMs; i.e. users across languages should get responses of similar quality irrespective of language used. In this work, we create a set of real-world open-ende...
FedUMM: A General Framework for Federated Learning with Unified Multimodal Models
arXiv:2601.15390v1 Announce Type: new
Abstract: Unified multimodal models (UMMs) are emerging as strong foundation models that can do both generation and understanding tasks in a single architecture. However, they are typically trained in centralized settings where all training and downstream datas...
Call2Instruct: Automated Pipeline for Generating Q&A Datasets from Call Center Recordings for LLM Fine-Tuning
arXiv:2601.14263v1 Announce Type: new
Abstract: The adaptation of Large-Scale Language Models (LLMs) to specific domains depends on high-quality fine-tuning datasets, particularly in instructional format (e.g., Question-Answer - Q&A). However, generating these datasets, particularly from unstructur...