List your company

Question Answering on PopQA (Score)

52.7ScoreLatentMem
11.20421.97732.7543.523Dec 15, 2025Jan 13, 2026Feb 11, 2026Mar 12, 2026Apr 10, 2026May 9, 2026Jun 7, 2026

Evaluation results

MethodLinks
LatentMemMAS=AutoGen, Atom.=–, Coord.=✓, T-free=–, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0652.7
ConMemMAS=AutoGen, Atom.=✓, Coord.=✓, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0649.9
ReMe (dynamic)MAS=AutoGen, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0649.1
LatentMemMAS=CAMEL, Atom.=–, Coord.=✓, T-free=–, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0647.23
SimpleMemMAS=AutoGen, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0646.9
ReMe (dynamic)MAS=MacNet, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0646.36
ConMemMAS=CAMEL, Atom.=✓, Coord.=✓, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0645.73
ReMe (dynamic)MAS=CAMEL, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0645.33
ConMemMAS=MacNet, Atom.=✓, Coord.=✓, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0645.29
SimpleMemMAS=MacNet, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0644.97
LatentMemMAS=MacNet, Atom.=–, Coord.=✓, T-free=–, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0644.14
CDKC (w/ 2 round)Backbone=Llama-3.1-8B-Instruct, Phase=Knowledge Calibration2026.0243.93
G-MemoryMAS=MacNet, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0643.88
RAGBackbone=Qwen2.5-7B-Instruct, Phase=Fundamental Capabilities2026.0243.44
GenerativeMAS=MacNet, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0643.39
JoyAgentMAS=MacNet, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0643.22
RAGBackbone=Llama-3.1-8B-Instruct, Phase=Fundamental Capabilities2026.0242.73
G-MemoryMAS=AutoGen, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0642.67
SimpleMemMAS=CAMEL, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0642.26
OAgentsMAS=MacNet, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0641.9
JoyAgentMAS=AutoGen, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0641.89
OAgentsMAS=AutoGen, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0640.7
GenerativeMAS=AutoGen, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0640.37
CDKCBackbone=Llama-3.1-8B-Instruct, Phase=Knowledge Calibration2026.0240.33
No-memoryMAS=AutoGen, Atom.=–, Coord.=–, T-free=–, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0638.78
G-MemoryMAS=CAMEL, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0638.04
JoyAgentMAS=CAMEL, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0637.5
MetaGPTMAS=CAMEL, Atom.=–, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0637.47
GRPOBackbone=Llama-3.1-8B-Instruct, Phase=Knowledge Calibration2026.0237.33
Olmo 3 Think (DPO)Training Stage=DPO, Model Family=Olmo 3, Parameter Count=32B, Thinking Capability=true2025.1237
ChatDevMAS=CAMEL, Atom.=–, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0636.78
VoyagerMAS=CAMEL, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0636.25
VoyagerMAS=MacNet, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0635.38
ChatDevMAS=MacNet, Atom.=–, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0635.33
GenerativeMAS=CAMEL, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0635.32
BARRELBackbone=Llama-3.1-8B-Instruct, Phase=Knowledge Calibration2026.0235
MetaGPTMAS=MacNet, Atom.=–, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0633.98
MetaGPTMAS=AutoGen, Atom.=–, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0633.8
VoyagerMAS=AutoGen, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0633.56
ChatDevMAS=AutoGen, Atom.=–, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0633.24
Olmo 3 Think (SFT)Training Stage=SFT, Model Family=Olmo 3, Parameter Count=32B, Thinking Capability=true2025.1233.1
No-memoryMAS=CAMEL, Atom.=–, Coord.=–, T-free=–, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0632.38
Qwen 3 VL 32B ThinkModel Family=Qwen 3, Parameter Count=32B, Thinking Capability=true2025.1232.2
K2-V2 70B InstructModel Family=K2, Parameter Count=70B, Thinking Capability=false2025.1232.2
OAgentsMAS=CAMEL, Atom.=✓, Coord.=–, T-free=✓, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0631.99
Olmo 3 Think (Final 3.0)Training Stage=Final Think 3.0, Model Family=Olmo 3, Parameter Count=32B, Thinking Capability=true2025.1231.9
CDKC (w/ 2 round)Backbone=Qwen2.5-7B-Instruct, Phase=Knowledge Calibration2026.0231.67
Olmo 3.1 Think 32BTraining Stage=Final Think 3.1, Model Family=Olmo 3.1, Parameter Count=32B, Thinking Capability=true2025.1230.9
Qwen 3 32BModel Family=Qwen 3, Parameter Count=32B, Thinking Capability=false2025.1230.7
Vanilla SFTBackbone=Llama-3.1-8B-Instruct, Phase=Knowledge Expansion2026.0229.43
Qwen 3 VL 8B Think2025.1229.3
CRew-DPOBackbone=Llama-3.1-8B-Instruct, Phase=Knowledge Calibration2026.0228.97
Granite 3.3 8B Inststage=Instruct2025.1228.9
CDKCBackbone=Qwen2.5-7B-Instruct, Phase=Knowledge Calibration2026.0228.27
GRPOBackbone=Qwen2.5-7B-Instruct, Phase=Knowledge Calibration2026.0228.1
Know WhatBackbone=Llama-3.1-8B-Instruct, Phase=Knowledge Calibration2026.0226.97
DS-R1 32BModel Family=DeepSeek-R1, Parameter Count=32B, Thinking Capability=true2025.1226.7
CGKEBackbone=Llama-3.1-8B-Instruct, Phase=Knowledge Expansion2026.0226.6
Qwen 3 VL 8B Inststage=Instruct2025.1226.5
OLMo 2 7B Inststage=Instruct2025.1225.5
Apertus 8B Inststage=Instruct2025.1225.5
LLKD-SFTBackbone=Llama-3.1-8B-Instruct, Phase=Knowledge Expansion2026.0225.07
No-memoryMAS=MacNet, Atom.=–, Coord.=–, T-free=–, Backbone=Qwen/Qwen3-4B-Instruct-25072026.0624.89
Olmo 3 7B ThinkStage=DPO2025.1224.7
CoTBackbone=Llama-3.1-8B-Instruct, Phase=Fundamental Capabilities2026.0224.33
Qwen 3 8B2025.1224.3
Olmo 3 7B ThinkStage=Final Think2025.1223.7
Qwen 2.5 7Bstage=Instruct2025.1221.5
BARRELBackbone=Qwen2.5-7B-Instruct, Phase=Knowledge Calibration2026.0220.97
Olmo 3 7B ThinkStage=SFT2025.1220.8
Olmo 3 7B Instructstage=DPO2025.1220.7
Qwen 3 8Bstage=Instruct2025.1220.4
Vanilla LLMBackbone=Llama-3.1-8B-Instruct, Phase=Fundamental Capabilities2026.0220.33
OpenThinker3 7B2025.1218
Nemotron Nano 9B v22025.1217.9
CRew-DPOBackbone=Qwen2.5-7B-Instruct, Phase=Knowledge Calibration2026.0216.57
Know WhatBackbone=Qwen2.5-7B-Instruct, Phase=Knowledge Calibration2026.0216.53
Olmo 3 7B Instructstage=SFT2025.1216.5
CGKEBackbone=Qwen2.5-7B-Instruct, Phase=Knowledge Expansion2026.0216.13
Vanilla SFTBackbone=Qwen2.5-7B-Instruct, Phase=Knowledge Expansion2026.0215.9
LLKD-SFTBackbone=Qwen2.5-7B-Instruct, Phase=Knowledge Expansion2026.0215.43
CoTBackbone=Qwen2.5-7B-Instruct, Phase=Fundamental Capabilities2026.0215.07
OR Nemotron 7B2025.1214.5
Vanilla LLMBackbone=Qwen2.5-7B-Instruct, Phase=Fundamental Capabilities2026.0214.5
Olmo 3 7B Instructstage=Final Instruct2025.1214.1
DS-R1 Qwen 7B2025.1212.8