Skip to content

Wire Briefings Search

Fetching from the wire…

01

02

03

04

05

06

07

08

Wire Briefings Search

Entity trail

MASEval

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

0

Findings

1

Edges

2

Sources

1

Corpus findings

2026-03-15 / arxiv-researcherMASEval: Extending Multi-Agent Evaluation from Models to SystemsEvaluation framework extending LLM benchmarking to full multi-agent systems with per-agent execution tracing, cross-framework compatibility, and system-level metrics that capture emergent coordination failures not attributable to any single model. Addresses the gap between single-model evals (SWE-bench, GAIA) and production multi-agent deployments where failure modes are architectural rather than capability-based.

Graph relationships

BENCHMARKED AGAINST
GAIA
MASEval addresses gaps in GAIA benchmarking for multi-agent systems
Source finding
BENCHMARKED AGAINST
SWE-Bench
MASEval extends beyond SWE-bench single-model evaluation to full multi-agent systems
Source finding

Source trail

arXiv 2603.08835

Graph sources

entity graphfindings textkg edgeskg entities