[KongchangAI]
PaperHierarchical Collaborative Memory Retrieval

HiCoMER

A hierarchical memory retrieval framework designed for multi-agent collaborative scenarios. It addresses the problem of semantically relevant but outdated memories being incorrectly retrieved by first assessing memory validity before executing retrieval. The framework comprises three components: a hierarchical memory conflict updater, a validity-aware retriever, and a memory-anchored answer generator.

Timeline (last 90 days)

Sep 28

HiCoMER is a memory-augmented framework for multi-agent collaborative scenarios, whose key design is to first maintain the validity of team memory and individual memory, and then retrieve from memories that are still valid

Unverified50%
Sep 28

The HiCoMER framework consists of three modules: a hierarchical memory conflict updater, a validity-aware retriever, and a memory-anchored answer generator

Unverified50%
Sep 28

The hierarchical memory conflict updater is responsible for maintaining the validity of team and individual memories, and handles conflicts between individual memories and team consensus

Unverified50%
Sep 28

The validity-aware retriever retrieves only from memories that are still valid, rather than performing full retrieval

Unverified50%
Sep 28

The hierarchical memory conflict updater manages memory by attaching validity flags to memory entries, marking individual memory entries that contradict new consensus as invalid rather than deleting them outright when the team reaches a new consensus

Unverified50%
Sep 28

HiCoMER decouples the two questions of 'what memories still hold' and 'which memory is most relevant,' solving them independently in separate stages

Unverified50%
Sep 28

HiCoMER regards team memory as representing the current collective consensus and thus having higher authority, while individual memory may contain intermediate states that have not yet converged or have already been overturned

Unverified50%
Sep 28

The research team constructed two new datasets oriented toward collaborative scenarios and based on memory-grounded QA to evaluate HiCoMER

Unverified50%
Sep 28

In experiments on the two datasets, HiCoMER maintains a consistent advantage over multiple strong baseline methods, specifically reflected in three aspects: significantly reducing retrieval of outdated memories, better preserving team consensus, and improving QA quality

Unverified50%
Sep 28

The two datasets of HiCoMER are designed to cover agent teams of different scales and different frequencies of memory updates

Unverified50%

1 more timeline events

All Facts (11)

Unverified

HiCoMER is a memory-augmented framework for multi-agent collaborative scenarios, whose key design is to first maintain the validity of team memory and individual memory, and then retrieve from memories that are still valid

50%
Unverified

The HiCoMER framework consists of three modules: a hierarchical memory conflict updater, a validity-aware retriever, and a memory-anchored answer generator

50%
Unverified

The hierarchical memory conflict updater is responsible for maintaining the validity of team and individual memories, and handles conflicts between individual memories and team consensus

50%
Unverified

The validity-aware retriever retrieves only from memories that are still valid, rather than performing full retrieval

50%
Unverified

The hierarchical memory conflict updater manages memory by attaching validity flags to memory entries, marking individual memory entries that contradict new consensus as invalid rather than deleting them outright when the team reaches a new consensus

50%
Unverified

HiCoMER decouples the two questions of 'what memories still hold' and 'which memory is most relevant,' solving them independently in separate stages

50%
Unverified

HiCoMER regards team memory as representing the current collective consensus and thus having higher authority, while individual memory may contain intermediate states that have not yet converged or have already been overturned

50%
Unverified

The research team constructed two new datasets oriented toward collaborative scenarios and based on memory-grounded QA to evaluate HiCoMER

50%
Unverified

In experiments on the two datasets, HiCoMER maintains a consistent advantage over multiple strong baseline methods, specifically reflected in three aspects: significantly reducing retrieval of outdated memories, better preserving team consensus, and improving QA quality

50%
Unverified

The two datasets of HiCoMER are designed to cover agent teams of different scales and different frequencies of memory updates

50%
Unverified

HiCoMER was proposed as an arXiv paper

50%

Source Articles