Related papers: SPICE: Self-Play In Corpus Environments Improves R…

Self-training systems often degenerate due to the lack of an external criterion for judging data quality, leading to reward hacking and semantic drift. This paper provides a proof-of-concept system architecture for stable self-training…

Recently, there has been a great deal of research in emergent communication on artificial agents interacting in simulated environments. Recent studies have revealed that, in general, emergent languages do not follow the compositionality…

In typical reinforcement learning (RL), the environment is assumed given and the goal of the learning is to identify an optimal policy for the agent taking actions through its interactions with the environment. In this paper, we extend this…

In this work, we introduce SPLICE, a human-curated benchmark derived from the COIN instructional video dataset, designed to probe event-based reasoning across multiple dimensions: temporal, causal, spatial, contextual, and general…

Motivation: Developing high-performing bioinformatics models typically requires repeated cycles of hypothesis formulation, architectural redesign, and empirical validation, making progress slow, labor-intensive, and difficult to reproduce.…

Can a model learn to escape its own learning plateau? Reinforcement learning methods for finetuning large reasoning models stall on datasets with low initial success rates, and thus little training signal. We investigate a fundamental…

Prompt optimization automatically refines prompting expressions, unlocking the full potential of LLMs in downstream tasks. However, current prompt optimization methods are costly to train and lack sufficient interpretability. This paper…

We introduce and evaluate Stated Preference for Interaction and Continued Engagement (SPICE), a simple diagnostic signal elicited by asking a Large Language Model a YES or NO question about its willingness to re-engage with a user's…

Reinforcement Learning has emerged as a strong alternative to solve optimization tasks efficiently. The use of these algorithms highly depends on the feedback signals provided by the environment in charge of informing about how good (or…

Large Language Models (LLMs) as agents often struggle in out-of-distribution (OOD) scenarios. Real-world environments are complex and dynamic, governed by task-specific rules and stochasticity, which makes it difficult for LLMs to ground…

Self-evolving large language models (LLMs) learn by generating their own training tasks and solutions, reducing reliance on human-curated supervision. However, in many reasoning domains, the model must also validate generated tasks and…

Text-to-image generation powers content creation across design, media, and data augmentation. Post-training of text-to-image generative models is a promising path to improve human preference alignment, factuality, and aesthetics. We…

Reasoning in a complex and ambiguous environment is a key goal for Reinforcement Learning (RL) agents. While some sophisticated RL agents can successfully solve difficult tasks, they require a large amount of training data and often…

We introduce Semantic Parsing in Contextual Environments (SPICE), a task designed to enhance artificial agents' contextual awareness by integrating multimodal inputs with prior contexts. SPICE goes beyond traditional semantic parsing by…

In this work, we extend the sparse iterative covariance-based estimator (SPICE), by generalizing the formulation to allow for different norm constraints on the signal and noise parameters in the covariance model. For a given norm, the…

A persistent challenge in AI is the effective integration of material and formal inference - the former concerning the plausibility and contextual relevance of arguments, while the latter focusing on their logical and structural validity.…

Large Language Models (LLMs) excel in many areas but continue to face challenges with complex reasoning tasks, such as Multi-Hop Question Answering (MHQA). MHQA requires integrating evidence from diverse sources while managing intricate…

We present GRACE, a simulation-native agent for autonomous experimental design in high-energy and nuclear physics. Given multimodal input in the form of a natural-language prompt or a published experimental paper, the agent extracts a…

Scaling the amount of compute used to train language models has dramatically improved their capabilities. However, when it comes to inference, we often limit models to making only one attempt at a problem. Here, we explore inference compute…

Despite recent advances, vision-language models trained with standard contrastive objectives still struggle with compositional reasoning -- the ability to understand structured relationships between visual and linguistic elements. This…