BabyAI
PulseAugur coverage of BabyAI — every cluster mentioning BabyAI across labs, papers, and developer communities, ranked by signal.
1 day(s) with sentiment data
-
ETHER agent improves reinforcement learning with emergent communication
Researchers have developed ETHER (Emergent Textual Hindsight Experience Replay), a novel agent designed to improve sample efficiency in goal-conditioned reinforcement learning (RL). ETHER addresses limitations of tradit…
-
New DDO method enhances LLM agent strategy diversity
Researchers have introduced Direct Diversity Optimization (DDO), a novel offline post-training method for large language model agents. DDO aims to improve the breadth of successful strategies an agent can employ by comb…
-
Prefix-GRPO enhances small language models for interactive agents
Researchers have introduced Prefix-GRPO, a novel reinforcement learning framework designed to enhance the performance of small language models in interactive agent tasks. This method decomposes teacher trajectories into…
-
AI world models suffer from instruction leakage, researchers find
Researchers have identified a critical issue in compact world models that use language goals to ground spatial relations, such as "put the red block left of the blue block." They found that these models often exhibit "i…