Cognitive Extensions for Dual-Process Language Agents
Cognitive Extensions for Dual-Process Language Agents: Memory and Self-Reflection in Interactive Environments
This academic paper details the design and benchmark results of cognitive extensions for language agents, providing insights into improving long-horizon decision-making in interactive environments.
Researchers extended SwiftSage, a dual-process AI agent, with an Adaptive Memory Module (AMM) and a Self-Reflection Module (SRM).
The full system, combining both extensions, achieved the best performance on ScienceWorld, with a mean final score of 64.
62, a success rate of 43.17%, and 19.
33 steps per successful action.The Self-Reflection Module was the strongest standalone contributor.
Cognitive Extensions for Dual-Process Language Agents: Memory and Self-Reflection in Interactive Environments
Language agents remain brittle in interactive environments, where success requires long-horizon state tracking, valid action execution, and recovery from failed steps. We extend SwiftSage, a dual-process agent that combines a fast action proposer with a slower planner, using two modular cognitive extensions: an Adaptive Memory Module (AMM) for salience-gated episodic storage and trigger-driven retrieval, and a Self-Reflection Module (SRM) for bounded execution-time validation and corrective intervention. Both modules are implemented as feature-flagged extensions over the same execution substrate, enabling controlled ablations on ScienceWorld. Across four configurations---baseline, baseline+AMM, baseline+SRM, and the full system---the full system achieves the best mean final score (64.62), success rate (43.17%), and successful-step efficiency (19.33 steps), while SRM is the strongest standalone contributor. The results suggest that execution-time control is the dominant bottleneck in this setting, while episodic memory becomes most useful once the runtime loop is stabilized.
This report is categorized under AITOP's "Models" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows and US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include SwiftSage, language agents, AI research, cognitive extensions.
Source attribution: curated from arXiv cs.AI (@João Meneses dos Santos, Arlindo L. Oliveira).
