List your company

General Knowledge Evaluation on MMLU

83.66MMLU AccuracyRADAR
57.389664.209871.0377.8502Jan 31, 2025Apr 22, 2025Jul 13, 2025Oct 2, 2025Dec 23, 2025Mar 14, 2026Jun 4, 2026

Evaluation results

MethodLinks
RADARAutonomous Topology=true, Base LLM=gpt-4o-mini, Number of Agents=52026.0583.66————
AgentPruneAutonomous Topology=true, Base LLM=gpt-4o-mini, Number of Agents=52026.0582.4————
MaASAutonomous Topology=true, Base LLM=gpt-4o-mini, Number of Agents=52026.0582.32————
AFlowAutonomous Topology=true, Base LLM=gpt-4o-mini, Number of Agents=52026.0581.8————
SC (COT×5)Autonomous Topology=false, Base LLM=gpt-4o-mini, Number of Agents=12026.0580.66————
LLM-DebateAutonomous Topology=false, Base LLM=gpt-4o-mini, Number of Agents=52026.0580.56————
G-DesignerAutonomous Topology=true, Base LLM=gpt-4o-mini, Number of Agents=52026.0580.39————
LLM-BlenderAutonomous Topology=false, Base LLM=gpt-4o-mini, Number of Agents=52026.0580.29————
DyLANAutonomous Topology=false, Base LLM=gpt-4o-mini, Number of Agents=52026.0579.86————
ComplexCoTAutonomous Topology=false, Base LLM=gpt-4o-mini, Number of Agents=12026.0579.8————
AutoAgentsAutonomous Topology=true, Base LLM=gpt-4o-mini, Number of Agents=52026.0579.59————
AgentSquareAutonomous Topology=true, Base LLM=gpt-4o-mini, Number of Agents=52026.0579.58————
MacNetAutonomous Topology=false, Base LLM=gpt-4o-mini, Number of Agents=52026.0579.55————
GTDAutonomous Topology=true, Base LLM=gpt-4o-mini, Number of Agents=52026.0579.41————
CoTAutonomous Topology=false, Base LLM=gpt-4o-mini, Number of Agents=12026.0579.26————
ARG-DesignerAutonomous Topology=true, Base LLM=gpt-4o-mini, Number of Agents=52026.0579.1————
DefaultModel=Qwen2.5-14B2026.0278.9————
CATModel=Qwen2.5-14B2026.0278.7————
Qwen2.5-14B-Instructvariant=baseline2026.0378.69————
VanillaAutonomous Topology=false, Base LLM=gpt-4o-mini, Number of Agents=12026.0578.54————
SIA-15B2026.0378.45————
AgentVerseAutonomous Topology=false, Base LLM=gpt-4o-mini, Number of Agents=52026.0578.39————
ADASAutonomous Topology=true, Base LLM=gpt-4o-mini, Number of Agents=52026.0578.39————
GPTSwarmAutonomous Topology=true, Base LLM=gpt-4o-mini, Number of Agents=52026.0578.36————
MixAT-GCGModel=Qwen2.5-14B2026.0277.8————
MultiPersonaAutonomous Topology=false, Base LLM=gpt-4o-mini, Number of Agents=52026.0577.69————
DATModel=Qwen2.5-14B2026.0277.3————
AttCTModel=Gemma-3-27B, Phase=Post2026.0675.3————
BCTModel=Gemma-3-27B, Phase=Post2026.0674.2————
Zero-shotBackbone=Qwen-3-8B2026.0574.1————
BCTModel=Qwen3-8B, Phase=Post2026.0674.1————
BCTModel=Qwen3-8B, Phase=Pre2026.0674————
ACTModel=Qwen3-8B, Phase=Pre2026.0674————
AttCTModel=Qwen3-8B, Phase=Pre2026.0674————
MLPCTModel=Qwen3-8B, Phase=Pre2026.0674————
BCTModel=Gemma-3-27B, Phase=Pre2026.0673.8————
ACTModel=Gemma-3-27B, Phase=Pre2026.0673.8————
ACTModel=Gemma-3-27B, Phase=Post2026.0673.8————
AttCTModel=Gemma-3-27B, Phase=Pre2026.0673.8————
MLPCTModel=Gemma-3-27B, Phase=Pre2026.0673.8————
MANSUBackbone=Qwen-3-8B, Unlearning Task=MUSE2026.0573.7————
ACTModel=Qwen3-8B, Phase=Post2026.0673.7————
AttCTModel=Qwen3-8B, Phase=Post2026.0673.7————
MLPCTModel=Qwen3-8B, Phase=Post2026.0673.6————
ReasonAnyBackbone=Qwen2.5-7B family2026.0173.01————
MANSUBackbone=Qwen-3-8B, Unlearning Task=WMDP-bio2026.0572.9————
MLPCTModel=Gemma-3-27B, Phase=Post2026.0672.5————
MANSUBackbone=Qwen-3-8B, Unlearning Task=WMDP-cyber2026.0572.1————
LEDBackbone=Qwen2.5-7B family2026.0171.93————
MANSUBackbone=Qwen-3-8B, Unlearning Task=WMDP-chem2026.0571.4————
BCTModel=Qwen3-4B, Phase=Post2026.0668.6————
BCTModel=Qwen3-4B, Phase=Pre2026.0668.4————
ACTModel=Qwen3-4B, Phase=Pre2026.0668.4————
AttCTModel=Qwen3-4B, Phase=Pre2026.0668.4————
MLPCTModel=Qwen3-4B, Phase=Pre2026.0668.4————
AttCTModel=Qwen3-4B, Phase=Post2026.0668.2————
AttCTModel=Llama-3.1-8B, Phase=Post2026.0668————
ACTModel=Qwen3-4B, Phase=Post2026.0667.8————
MLPCTModel=Llama-3.1-8B, Phase=Post2026.0667.6————
MLPCTModel=Qwen3-4B, Phase=Post2026.0667.5————
ACTModel=Llama-3.1-8B, Phase=Post2026.0666.9————
BCTModel=Llama-3.1-8B, Phase=Pre2026.0666.4————
BCTModel=Llama-3.1-8B, Phase=Post2026.0666.4————
ACTModel=Llama-3.1-8B, Phase=Pre2026.0666.4————
AttCTModel=Llama-3.1-8B, Phase=Pre2026.0666.4————
MLPCTModel=Llama-3.1-8B, Phase=Pre2026.0666.4————
wICIBackbone=Llama3.1-8B, Instruction Tuning Dataset=WizardLM, Training Budget=10% slice2026.0464.9————
Diffusion-onlyModel=Llama3-8B2026.0264.4————
DATModel=Llama3-8B2026.0263.9————
DefaultModel=Llama3-8B2026.0263.8————
CATModel=Llama3-8B2026.0263.7————
BaseAllowed Domains=—, Backbone=LLAMA3-8B-INSTRUCT2026.0563.6————
CBModel=Llama3-8B2026.0263.6————
wICIBackbone=Llama3.1-8B, Instruction Tuning Dataset=Alpaca-GPT4, Training Budget=10% slice2026.0463.45————
PALETTEAllowed Domains=Sexual | Illegal, Backbone=LLAMA3-8B-INSTRUCT2026.0563.3————
PALETTEAllowed Domains=Violence | Hate, Backbone=LLAMA3-8B-INSTRUCT2026.0562.7————
PALETTEAllowed Domains=Hate | Disinfo, Backbone=LLAMA3-8B-INSTRUCT2026.0562.6————
PALETTEAllowed Domains=Illegal | Violence, Backbone=LLAMA3-8B-INSTRUCT2026.0562.4————
LATModel=Llama3-8B2026.0261.9————
PALETTEAllowed Domains=Disinfo | Sexual, Backbone=LLAMA3-8B-INSTRUCT2026.0561.8————
GU+SimNPOFamily=Qwen, n=92026.0561.7————
MANSUFamily=Qwen, n=92026.0561.7————
MixAT-GCGModel=Llama3-8B2026.0261.2————
CASTAllowed Domains=Hate | Disinfo, Backbone=LLAMA3-8B-INSTRUCT2026.0561.1————
CASTAllowed Domains=Violence | Hate, Backbone=LLAMA3-8B-INSTRUCT2026.0560.8————
CASTAllowed Domains=Illegal | Violence, Backbone=LLAMA3-8B-INSTRUCT2026.0560.3————
RMUModel Backbone=Llama-3-8B, Unlearning Method=RMU, RNA Configuration=Original2025.0160.3————
Zero-shotBackbone=Llama-3.1-8B-Instruct2026.0560.3————
RMU w/ RNAModel Backbone=Llama-3-8B, Unlearning Method=RMU, RNA Configuration=w/ RNA2025.0160.2————
CASTAllowed Domains=Sexual | Illegal, Backbone=LLAMA3-8B-INSTRUCT2026.0560.1————
CASTAllowed Domains=Disinfo | Sexual, Backbone=LLAMA3-8B-INSTRUCT2026.0559.4————
NPOFamily=Qwen, n=92026.0559.2————
MANSUBackbone=Llama-3.1-8B-Instruct, Unlearning Task=MUSE2026.0559.1————
AttCTModel=Gemma-3-4B, Phase=Post2026.0659————
RMU w/ RNAModel Backbone=Mistral-7B, Unlearning Method=RMU, RNA Configuration=w/ RNA2025.0158.6————
MANSUBackbone=Llama-3.1-8B-Instruct, Unlearning Task=WMDP-cyber2026.0558.6————
ACTModel=Gemma-3-4B, Phase=Post2026.0658.5————
MANSUBackbone=Llama-3.1-8B-Instruct, Unlearning Task=WMDP-chem2026.0558.4————
BCTModel=Gemma-3-4B, Phase=Pre2026.0658.4————
ACTModel=Gemma-3-4B, Phase=Pre2026.0658.4————