bug-bounty529
xss292
rce171
google149
account-takeover120
exploit119
facebook118
bragging-post118
malware105
privilege-escalation104
microsoft101
open-source94
authentication-bypass90
csrf89
cve82
stored-xss75
access-control75
ai-agents65
web-security64
phishing63
reflected-xss63
writeup56
input-validation52
ssrf51
reverse-engineering51
cross-site-scripting50
sql-injection50
smart-contract49
information-disclosure49
defi48
apple48
tool47
api-security46
ethereum45
privacy44
vulnerability-disclosure43
browser39
opinion38
llm38
web-application38
burp-suite37
ai-security37
web337
race-condition36
automation36
remote-code-execution36
responsible-disclosure35
supply-chain35
dos35
lfi34
0
5/10
NVIDIA AI-Q achieved first place on DeepResearch Bench I and II by implementing a multi-agent architecture (orchestrator, planner, researcher) powered by fine-tuned Nemotron 3 Super models, trained on 67k SFT trajectories from research synthesis tasks, demonstrating that open, modular agentic systems can achieve state-of-the-art performance on research agent benchmarks.
ai-agents
llm
benchmark
research-synthesis
multi-agent-architecture
fine-tuning
tool-use
retrieval-augmented-generation
citation-management
NVIDIA AI-Q
DeepResearch Bench
DeepResearch Bench II
NVIDIA NeMo Agent Toolkit
NVIDIA Nemotron 3
LangChain DeepAgents
Nemotron-3-Super-120B-A12B
Tavily
Serper
OpenScholar
ResearchQA
Fathom-DeepResearch-SFT
GPT-OSS-120B
Qwen3-Nemotron-32B-GenRM-Principle