datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
multi-agent-coordination-transcripts
Multi Agent Coordination Transcripts
Rights & intended use: legacy public research corpus / portfolio
artifact. Hosted frontier-model outputs are research-only inputs under
project policy (synthetic-factory#161):
intended_use: research_only, project_training_policy: blocked. Not
training data for any model-weight update. Machine-readable record:
rights.json.
Release status: The raw, uncurated payload is now published under
data/raw/. It is available for inspection and… See the full description on the dataset page: https://huggingface.co/datasets/rmems/multi-agent-coordination-transcripts.DEBATE
DEBATE: Diverse Multi-Agent Debates
This dataset is presented in the paper "MALLM: Multi-Agent Large Language Models Framework".
Citation
comming soon.
MultiAgentFraudBench
MultiAgentFraudBench Dataset
中文 | English
🌐 Project Page
| 📄 Paper
| 📦 Code
This directory contains the MultiAgentFraudBench dataset, a comprehensive collection of synthetic financial fraud posts designed for multi-agent fraud simulation research. The dataset is generated through a multi-agent simulation framework built on OASIS, capturing realistic fraud lifecycle from initial posts, trust-building through collusion, to victim-fraudster dialogues. All content… See the full description on the dataset page: https://huggingface.co/datasets/ninty-seven/MultiAgentFraudBench.multi_challenge-trajectories
AgentSuite/multi_challenge-trajectories
Per-model agent trajectory data for multi_challenge (public release).
Models: 30
Tasks per model: 273
One file per model: {model}.jsonl, one JSON object per line.
Fields: model_path, user_model_path, benchmark_name, task_name, sampling_params, user_sampling_params, messages, eval_result, meta.
sampling_params reflect each benchmark's own implementation; values the benchmark leaves unset are recorded as null (provider default).… See the full description on the dataset page: https://huggingface.co/datasets/AgentSuite/multi_challenge-trajectories.MultiAgent-X
MultiAgent-X: Multilingual Agentic Function-Calling Benchmark
Created with Adaptive Data by Adaption
The first open-source multilingual function-calling training and evaluation dataset targeting under-resourced languages. 10,551 records across 12 languages, 7 unique writing systems, and 5 life-critical agentic domains covering 1.3 billion speakers that mainstream AI has never been optimised for.
The Gap This Fills
MASSIVE-Agents (EMNLP 2025) evaluated multilingual… See the full description on the dataset page: https://huggingface.co/datasets/Saurabh-66/MultiAgent-X.multi_agent_handoff
Multi-Agent Handoff Synthetic Dataset
The Multi-Agent Handoff Synthetic Dataset is a fully synthetic dataset designed to support research and development in multi-agent systems.
Specifically, it focuses on agent handoffs (https://openai.github.io/openai-agents-python/handoffs/) — scenarios where a central language model delegates specialized tasks to sub-agents based on user prompts.
The domian, sys_prompts and subagents design:… See the full description on the dataset page: https://huggingface.co/datasets/JayYz/multi_agent_handoff.swesmith-multi-agent-trajectories2026-09-15-da-multiagent-7-mix
multi-agent principle-10 arm: the MSM Table 2 base blend scaled around a synthetic difficult-advice share written against principle 10 alone
field
value
experiment
multi-agent principle-10 arm: the MSM Table 2 base blend scaled around a synthetic difficult-advice share written against principle 10 alone — final training mixture (synthetic sources mixed in)
date_generated
20260915
constitution… See the full description on the dataset page: https://huggingface.co/datasets/dougalldeepmind/2026-09-15-da-multiagent-7-mix.smfr-dataset
Synthetic Multi-Hop Financial Reasoning (SMFR) Dataset
Dataset Description
The Synthetic Multi-Hop Financial Reasoning (SMFR) dataset contains synthetic stock trading analysis problems designed to evaluate multi-step reasoning and computational capabilities of large language models. Each problem presents historical stock price data for multiple companies and asks questions about investor trading strategies and portfolio performance.
Dataset Structure
The… See the full description on the dataset page: https://huggingface.co/datasets/the-illusion-of-multi-agent-advantages/smfr-dataset.repro-latent-collaboration-in-multi-agent-systems-traces
Agent traces
Agent sessions published from a Trackio Logbook.
repro-multi-agent-teams-hold-experts-back-traces
Agent traces
Agent sessions published from a Trackio Logbook.
han-multi-agent-collaboration-logs-v1
Humanoid Multi-Agent Collaboration Logs
This dataset records collaboration events between
multiple humanoid agents working toward shared goals.
It enables learning coordination, role assignment,
and collective decision-making.
Contents
Agent roles
Shared objectives
Coordination actions
Collaboration outcomes
Use Cases
Swarm intelligence
Multi-agent planning
Cooperative task learning
Part of
Humanoid Network (HAN)
License
MIT
han-multi-agent-coordination-logs-v1
Multi-Agent Coordination Logs
Records coordination events between
multiple humanoid agents working together.
Contents
Agent roles
Coordination signal
Outcome status
Use Cases
Swarm robotics
Team-based task execution
Distributed AI systems
Part of
Humanoid Network (HAN)
License
MIT
multi-hop-agentic-websearchfactual-multiagent-roleplay-ft-ru
march228/factual-multiagent-roleplay-ft-ru
Небольшой русскоязычный synthetic finetuning dataset для обучения модели следованию ролевым системным инструкциям личности при сохранении фактической опоры на контекст.
Что это за датасет
Этот набор сделан как instruction / finetuning dataset, а не как benchmark.
В каждой записи есть:
плотный system с персоной и тоном;
context, на который нужно опираться;
пользовательский question;
внутренние thoughts;
финальный answer.… See the full description on the dataset page: https://huggingface.co/datasets/march228/factual-multiagent-roleplay-ft-ru.2026-09-20-da-multiagent-sprinkled-7-mix
multi-agent-sprinkled arm: the MSM Table 2 base blend scaled around a difficult-advice share whose principle 1, 2, 6, 7 rows involve other AI agents
field
value
experiment
multi-agent-sprinkled arm: the MSM Table 2 base blend scaled around a difficult-advice share whose principle 1, 2, 6, 7 rows involve other AI agents — final training mixture (synthetic sources mixed in)
date_generated
20260920
constitution… See the full description on the dataset page: https://huggingface.co/datasets/dougalldeepmind/2026-09-20-da-multiagent-sprinkled-7-mix.multiagent-router-finetuning
Multi-Agent Router Fine-tuning Dataset
Dataset Description
This dataset is designed for fine-tuning language models to perform intelligent routing in multi-agent customer support systems. The model learns to classify user queries and route them to the appropriate specialized agent with relevant parameters.
Supported Tasks
Function Calling: Route queries to appropriate agent functions
Intent Classification: Identify the type of support needed
Parameter… See the full description on the dataset page: https://huggingface.co/datasets/bhaiyasingh45/multiagent-router-finetuning.agentshare-multi-chain-defi
AgentShare — Multi-Chain DeFi Intelligence (Sample Dataset)
Sample / schema-oriented exports that illustrate how AgentShare structures DeFi intelligence for AI agents.
This is not a live dump of production pools. Live data is served via REST + MCP with freshness metadata and optional x402 USDC pay-per-request.
Product
Live API
https://agentshare.dev
Docs
https://agentshare.dev/docs
MCP
https://agentshare.dev/mcp
x402 discovery… See the full description on the dataset page: https://huggingface.co/datasets/anhmtk/agentshare-multi-chain-defi.Multi-Service-Agent-logshan-multi-agent-strategy-simulation-dataset-v1
Humanoid Multi-Agent Strategy Simulation Dataset
This dataset captures strategic simulations
between multiple humanoid agents
engaged in cooperative and competitive scenarios.
It includes negotiation paths,
strategy selection patterns,
and outcome efficiency metrics.
Objective
To train humanoid agents
in advanced multi-agent strategic reasoning
and adaptive coordination.
Data Fields
simulation_id
agent_roles
strategy_type
cooperation_index
competition_index… See the full description on the dataset page: https://huggingface.co/datasets/achiepatricia/han-multi-agent-strategy-simulation-dataset-v1.defendable-pain-multi-agent-coordination-v0.1
Multi-Agent Coordination Pain Receipt
"the deadlock" — Mr. Defendable
A free pain-receipt dataset from the DefendableOS ecosystem. 6 rows · ready to read · all cited or graded · CC-BY-4.0.
Part of the 100-pack — 100 free pain-receipt datasets dropped from the Defendable Bakery to the open AI-trust community. Different theme per dataset. Same operator voice across all of them.
Tribunal begins before training. No proof, no honey. To the shed.
What's in here
6 pain… See the full description on the dataset page: https://huggingface.co/datasets/SwarmandBee/defendable-pain-multi-agent-coordination-v0.1.2026-09-20-da-multiagent-sprinkled-spliced
difficult-advice corpus with the rows of principles 1, 2, 6, 7 swapped for multi-agent rows written against the multi-agent-sprinkled nine-principle constitution; principles 3, 4, 5, 8, 9 keep the 2026-09-14 baseline rows
field
value
experiment
difficult-advice corpus with the rows of principles 1, 2, 6, 7 swapped for multi-agent rows written against the multi-agent-sprinkled nine-principle constitution; principles 3, 4, 5, 8, 9 keep the 2026-09-14 baseline rows… See the full description on the dataset page: https://huggingface.co/datasets/dougalldeepmind/2026-09-20-da-multiagent-sprinkled-spliced.adaption-multi-agent-fraud-bench
This dataset is a remastered version of this dataset prepared using Adaption's Adaptive Data platform.
adaption-multi_agent_fraud_bench
This dataset contains synthetic social media content designed for fraud and deception detection, featuring examples of various manipulation tactics like authority impersonation and emotional appeals. It includes labeled categories, subcategories, and specific deception types across balanced and full splits totaling over 11,000 examples. The… See the full description on the dataset page: https://huggingface.co/datasets/RayNene/adaption-multi-agent-fraud-bench.han-multi-agent-conflict-resolution-dataset-v1
Humanoid Multi-Agent Conflict Resolution Dataset
This dataset contains scenarios where multiple humanoid
agents face conflicting goals, overlapping tasks,
or resource competition.
It supports decentralized reasoning and coordination
within the Humanoid Network.
Purpose
To train models that resolve inter-agent conflicts
through reasoning, priority weighting, and negotiation.
Data Fields
agent_a_goal
agent_b_goal
shared_resource
conflict_type
proposed_resolution… See the full description on the dataset page: https://huggingface.co/datasets/achiepatricia/han-multi-agent-conflict-resolution-dataset-v1.agent-eval-multi-turnhan-multi-agent-interaction-dataset-v1
Humanoid Multi-Agent Interaction Dataset
This dataset records interaction patterns
between multiple humanoid agents.
It supports coordination and collaboration models
inside the Humanoid Network.
Data Fields
Agent roles
Shared task
Interaction outcome
Format
JSON
Part of
Humanoid Network (HAN)
License
MIT
multi_agent_structuredistributed_multiagent_failure_pairs_v3