Atlas Graph Explorer

Wiki Graph Edges

Clusters

catalog-meta1769

CiSurface1
Claim825
ClaimTest5
ClaimTestRun5
DeploymentTarget2
EvidencePolicy2
EvidenceSource227
Gap4
Grammar1
MetaAttribute366
MetaCluster16
MetaEdgeKind84
MetaEnum34
MetaNodeKind114
OntologySchema1
OpenQuestion5
OutOfScopeReason4
PackageSurface23
PathDescriptor45
RunJournalEvent5

extensions1288

domain1150

Domain108
Framework36
Language39
Library63
Platform10
PlatformService65
Responsibility76
Role96
SkillArea339
Specialization39
StackPart52
StackProfile7
Tool101
Topic179
Workflow691

workflows659

Workflow691

terminology444

Acronym25
Definition116
Synonym11
Term292

agent-stack349

AgentCoreImpl28
AgentPlatformImpl20
AgentProduct38
AgentRuntimeImpl24
AgentUIImpl18
AgentVersion42
CapabilityProfile3
InteractionPrimitive257
InteractionPrimitiveCategory12
LaunchConfig10
Presentation22
SessionModel4

benchmarks232

Benchmark65
EvalHarness5
EvalResult73
EvalRun70
Judge3
Rubric3
TestSet13

sourceref-scope175

ScopeBoundary111
SourceRef64

role163

Customer3
EndUser3
OrgUnit59
Responsibility76
Role96
Tenant3
VendorRelationship3

channels-hooks137

Channel6
DecisionVerb5
HookMapping54
HookSurface70
MergePolicy2

capabilities-and-models133

CapabilitySupport100
Modality7
ModelProviderProduct7
ModelProviderVersion7
ProcessDescriptor2
TransportProtocol10

compute116

APIErrorClass3
AgentHostTransport7
CapacityCascadeSignal2
MCPTransport4
ModelFamily23
ModelTransportProtocol12
ModelVersion49
Provider12
TranscriptIngressEndpoint1
TransportClient2
TransportProxy1

lifecycle106

capabilities85

Capability74
InstallMethod11

wiki44

Page44

security21

FilesystemSafetyInvariant3
HarnessHardeningGuidance3
OutputGuard5
OutputModeChange2
OutputStyle2
PermissionDenialReason3
SecretHandlingPolicy3

stack-layers11

Layer11

context-engineering8

BackgroundConsolidation2
CompactionPolicy2
MemoryHierarchy2
ProactiveSurface2

trust5

TrustLevel5

HomebenchmarksBenchmark

Benchmark

2 of 65 records

Active filters

kind: knowledge x targetsKind: ModelVersion x clear all

homepageUrl

https://www.swebench.com/2
https://github.com/THUDM/AgentBench1
https://github.com/hendrycks/apps1
https://os-world.github.io/1
https://google-research.github.io/android_world/1
https://metr.org/AI_R_D_Evaluation_Report.pdf1
https://appworld.dev/1
https://assistantbench.github.io/1
https://the-agent-company.com/1
https://agentclinic.github.io/1
https://osu-nlp-group.github.io/TravelPlanner/1
https://openai.com/index/browsecomp/1

kind

model-only14
code-generation7
full-stack7
web-agent7
reasoning5
math4
tool-use3
domain-specific2
agent-leaderboard2
knowledge2
research-engineering1
planning1

description

General AI Assistants benchmark — real-world agent reasoning tasks. 1
Hand-written programming problems for evaluating code generation. 1
MBPP+ from EvalPlus — augmented MBPP with substantially expanded test suites. 1
Machine learning engineering tasks drawn from Kaggle competitions. 1
Massive Multitask Language Understanding — 57-subject knowledge benchmark. 1
Real-world software engineering issues from open-source Python repos. 1

targetsKind

ModelVersion37
AgentVersion28

Sort:id-asc id-desc name-asc name-desc

id	displayName	cluster
benchmark:hle	Humanity's Last Exam (HLE)	benchmarks
benchmark:mmlu	MMLU	benchmarks