ASK+ trajectory-aware prompts lift RL agent success on FourRooms from 53% to 70%
Which summary reads better? Pick one — models revealed after.Both summaries are AI-generated.
By providing a small language model with trajectory-aware context and structured chain-of-thought, you can deploy a 2B-parameter model like Qwen3.5-2B to correct physical RL agents in partially observable environments, outperforming vanilla gating methods to reach up to a 70% success rate on task-solving like FourRooms. This proves that predictive entropy remains a viable gating signal for selective querying in POMDPs, and that prompt statefulness, rather than model scale, dictates the viability of low-latency, low-cost SLMs acting as real-time policy correctors. This allows you to deploy high-accuracy hybrid agent architectures in production without the latency and cost of larger 4B+ models.
Adding trajectory-aware context and structured reasoning to uncertainty-gated SLM assistance boosts performance by 17–40% in partially observable environments, proving prompt design outweighs model scale. This lets engineers deploy smaller, cheaper models effectively by focusing on context-rich prompts and selective querying instead of brute-force scaling.