AI Paper Atlas

Research map / Multi-agent systems and games / LLM agents

DreamPhase: Offline Imagination and Uncertainty-Guided Planning for Large-Language-Model Agents

Shayan Mohajer Hamidi, Linfeng Ye, Konstantinos N. Plataniotis

ICLR 2026 · Multi-agent systems and games › LLM agents · approach: Agent data synthesis and behavior studies

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes DreamPhase, an offline imagination and uncertainty-guided planning framework for LLM agents to improve closed-loop decision-making.

Problem

LLM agents struggle with closed-loop decision-making in multi-step tasks due to static pretraining and limited temporal grounding.

Approach

Uses offline imagination (simulated future rollouts) combined with uncertainty-guided planning so LLM agents can plan over imagined trajectories; the abstract is truncated so specifics are unavailable.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
plan with offline latent imagination plus uncertainty gate instead of ARMAP-M token searchARMAP-M, Llama-2-7B · Eight agent tasks averagesuccess rate avg: 42.3 -> 50.1Improved
use DreamPhase instead of ARMAP-MARMAP-M, Llama3-70B · WebShop/SciWorld/Game-of-24 averagesuccess rate avg: 51.2 -> 53.6Improved
use DreamPhase instead of ARMAP-MARMAP-M, Mistral-7B · WebShop/SciWorld/Game-of-24 averagesuccess rate avg: 28.9 -> 30.7Improved
use DreamPhase instead of ARMAP-MARMAP-M, Llama-8B · WebShopAPI calls per episode: 39.8 -> 9.3Improved
use DreamPhase instead of ARMAP-MARMAP-M, Llama-8B · WebShopsuccess rate: 60.2 -> 61.8Improved
add uncertainty safety gate and imaginationARMAP-M, Llama-8B · WebShopcritical irreversible incidents per 1000: 7.1 -> 1.4Improved
use DreamPhase instead of ARMAP-MARMAP-M, Llama-8B · ALFWorld-stdsuccess rate: 0.31 -> 0.39Improved
use DreamPhase instead of AgentGymAgentGym, Llama-2-7B · BabyAIsuccess rate: 82.7 -> 82.3No clear effect
use DreamPhase instead of ARMAP-MARMAP-M, Llama-2-7B · MAZE/BIRD taskssuccess rate: MAZE 12.0 -> 14.0; BIRD 10.5 -> 13.5Mixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy