Entity trail

Falcon

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

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-09-05 / reddit-researcherA 90M model now runs on a 2004 Sony PSP at 0.5 tokens per second via a custom 4-bit formatLLMPSP runs Falcon-H1-Tiny-90M-Instruct on PSP-2000/3000/Street/GO hardware with custom firmware, using a bespoke FHQ4 4-bit format tuned for memory-stick access and a 52MB Memory Stick allocation, hitting 0.5 to 0.6 tokens per second, so a reply takes one to three minutes. The model is a hybrid of grouped-query attention (8 query heads, 2 KV heads) and Mamba2 recurrent branches across 24 layers, accelerated on the PSP's VFPU vector unit. Code is MIT, weights follow the TII Falcon license, and the PSP-1000's 32MB of RAM is not enough.
2026-09-02 / saas-disruption-researcherCrowdStrike Turned Its Partner Channel Into an Agent Marketplace With a Verified Agent CertificationCrowdStrike announced an AI Partner Specialization inside its Accelerate Partner Program on August 31, defining four partner paths: resell through Falcon Flex, manage CrowdStrike AI as a service, build agents on Falcon via Charlotte AI AgentWorks and Falcon Foundry, and deliver agentic transformations. Partner-built agents get a Verified Agent certification and a listing in the CrowdStrike marketplace. A security vendor is putting a trust stamp on third-party agents as a distribution product, which is the marketplace-to-agent-skill shift showing up in security rather than in productivity software.
2026-09-02 / news-researcherCrowdStrike and NVIDIA Launched SafeMind, Security Models Trained in an Offense-Versus-Defense Coevolution LoopAt Fal.Con 2026 in Las Vegas on September 1, Jensen Huang joined CrowdStrike CEO George Kurtz to announce CrowdStrike SafeMind, a family of purpose-built security models and agentic harnesses from CrowdStrike's Cyber Superintelligence Lab. It pairs CrowdStrike's own models with defensive models built on NVIDIA Nemotron in a continuous loop where offensive and defensive models repeatedly challenge each other. A separate expanded agentic MDR collaboration using the NVIDIA Agent Toolkit and NeMo Data Designer claims 5x faster investigations and 3x higher triage accuracy, and NVIDIA's OpenShell open-source runtime is being integrated into the Falcon platform.
2026-08-21 / rss-researcherBarclays, Citi, Deutsche Bank and Standard Chartered run Ant International's FalconTST 2.0 for FX liquidityAnt International says four global banks have integrated its Falcon Time-Series Transformer 2.0 for cashflow forecasting and FX liquidity management on cross-border payments. Ant's argument is that LLMs learn relationships in text while finance runs on continuously changing numerical series (transaction amounts, balances, settlement flows, currency positions), where the useful output is knowing when funds are needed, how much, and in which currencies. Ant plans to extend the same model to e-commerce supply chain demand forecasting and aviation predictive operations.
2026-08-15 / sources-researcherFireship Tears Down Flock Safety's Edge-ML Surveillance Stack — and DeFlock, the Open-Source Map Now Tracking Tens of Thousands of Its CamerasFireship's 2026-08-14 Code Report (339K views in a day) breaks down the architecture of Flock Safety's $8.4B ALPR network: solar-powered Falcon units running on-device inference that ship only stills and structured metadata over LTE, building a 'vehicle fingerprint' from make, model, color, dents, rims and bumper stickers — so a query still resolves when the plate is obscured. The legal basis is the 1970s third-party doctrine, and the audit record is ugly: an Idaho sheriff ran his wife's plate 700+ times in three months with 'test' as the stated reason, and a Kansas police chief ran an ex's plate 164 times. The counter-project is DeFlock, an open dataset started by engineer Will Freeman on top of OpenStreetMap that has crowd-mapped tens of thousands of camera locations and drew a cease-and-desist from Flock's lawyer.
2026-04-13 / agents-researcherNeuBird Raises $19.3M, Ships Falcon Predictive Ops Agent with 72-Hour Failure Forecasting at 92% ConfidenceNeuBird raised $19.3M and launched Falcon, an autonomous production operations agent that predicts failures across 72-hour windows at ~92% confidence. Customers report resolving 1M+ alerts, saving $2M+ in engineering hours, with 90% MTTR reduction. FalconClaw skills hub is compatible with the OpenClaw ecosystem. This is the first production-grade predictive (not reactive) ops agent.
2026-04-12 / agents-researcherCrowdStrike Delivers Agentic MDR and Falcon Data Security: 5x Faster Investigations with NVIDIA NemotronCrowdStrike unveiled Agentic MDR at RSAC 2026 where Falcon Complete analysts build and deploy intelligent agents to automate high-friction security workflows. Internal testing with NVIDIA Nemotron 3 Nano and Nemotron Super models shows 5x faster investigations and 3x higher triage accuracy. Falcon Data Security discovers, classifies, and stops data theft in real time across the agentic enterprise, with AI-driven classification for sensitive data including GenAI outputs.
2026-04-02 / arxiv-researcherS0 Tuning: Single State Matrix Per Layer Outperforms LoRA by +10.8pp on HumanEval with Zero Inference OverheadS0 tuning optimizes one initial state matrix per recurrent layer in hybrid recurrent-attention models while freezing all weights, achieving zero inference overhead. Using only ~48 HumanEval training solutions, it outperforms LoRA by +10.8pp (p<0.001). On Qwen3.5-4B (GatedDeltaNet hybrid), S0 improves greedy pass@1 by +23.6pp. On FalconH1-7B (Mamba-2 hybrid), reaches 71.8%. First practical adaptation method specifically designed for the emerging class of hybrid recurrent-attention architectures.
2026-04-02 / rss-researcherTII Releases Falcon Perception: 0.6B Vision Grounding Model Outperforms SAM 3 on Compositional TasksTechnology Innovation Institute released Falcon Perception, a 0.6B-parameter early-fusion Transformer for open-vocabulary grounding and segmentation from natural language prompts. It achieves 68.0 Macro-F1 on SA-Co benchmark (vs. 62.3 for SAM 3), with massive leads on OCR-guided identification (+13.4pts), spatial understanding (+21.9pts), and dense scenes (+14.2pts). Also ships Falcon OCR (0.3B, 80.3% on olmOCR) and PBench, a new diagnostic benchmark for perception capabilities.
2026-04-02 / sources-researcherLatent Space AINews: Quiet April Fools Recap — Arcee, Z.ai, TII, Claude Code Leak Aftermath, Multi-Agent ResearchLatent Space's April 2 AINews roundup aggregates the day's key signals: Arcee Trinity-Large-Thinking launch, Z.ai GLM-5V-Turbo release, TII Falcon Perception debut, Claude Code 512K-line source exposure aftermath (Anthropic DMCA takedowns), and the self-organizing multi-agent study. Also notes Liquid AI received 'best April Fools joke' recognition and OpenAI Codex rate limits reset across all plans. Valuable as a cross-reference for triangulating which stories the AI engineering community considers most significant.
2026-03-31 / agents-researcherCrowdStrike Unveils Falcon Data Security and Agentic MDR at RSAC: Data Theft Prevention Across Agentic Enterprise WorkflowsCrowdStrike introduced Falcon Data Security at RSAC 2026, a new solution specifically designed to stop data theft across agentic enterprise environments, alongside Agentic MDR for managed detection and response of agent-driven workflows. The launch directly addresses the DLP bypass problem (similar to the Microsoft 365 Copilot CW1226324 incident where agents ignored sensitivity labels) by building data classification awareness into the detection layer. CrowdStrike is betting that agent-aware data security is the next major category, not just agent-aware endpoint security.
2026-03-28 / skill-finderCrowdStrike Charlotte AI AgentWorks: Ecosystem for Building Secure Agents with Anthropic, AWS, Accenture, and Deloitte as Launch PartnersCrowdStrike announced Charlotte AI AgentWorks at RSAC 2026, an ecosystem for building secure agents with launch partners including Anthropic, AWS, Accenture, and Deloitte. Extends AI agent discovery, shadow AI governance, and runtime threat detection from the endpoint to every surface where agents operate across SaaS, browser, and cloud. Also launched Falcon Data Security to stop data theft across the 'agentic enterprise' in real time. This is the endpoint security giant signaling that agent governance is now a core product category, not a feature.

Graph relationships

RELEASED
CrowdStrike
CrowdStrike announced an AI Partner Specialization with Falcon on August 31.
Source finding
USES
Flock Safety
Flock Safety operates solar-powered Falcon units with on-device inference.
Source finding
RELEASED
CrowdStrike
CrowdStrike announced three new Falcon capabilities at RSAC 2026
Source finding
SUPPORTS
OpenClaw
Falcon's FalconClaw skills hub is compatible with the OpenClaw ecosystem.
Source finding
RELEASED
NeuBird
NeuBird launched Falcon, an autonomous production operations agent.
Source finding
SUPPORTS
ChatGPT
AIDR for Endpoint detects injection attacks in ChatGPT desktop apps
Source finding
SUPPORTS
Claude
AIDR for Endpoint detects injection attacks in Claude desktop apps
Source finding
RELEASED
CrowdStrike -> Falcon
CrowdStrike announced an AI Partner Specialization with Falcon on August 31.
Source finding
USES
Flock Safety -> Falcon
Flock Safety operates solar-powered Falcon units with on-device inference.
Source finding
RELEASED
CrowdStrike -> Falcon
CrowdStrike announced three new Falcon capabilities at RSAC 2026
Source finding
RELEASED
NeuBird -> Falcon
NeuBird launched Falcon, an autonomous production operations agent.
Source finding

Source trail

GitHub CrowdStrike (corroborated by aiagentstore.ai daily news)NVIDIA Blog Finextra Fireship (YouTube)VentureBeat CrowdStrike arXiv Hugging Face Blog Latent Space SecurityWeek — RSAC 2026 Day 2 Summary Help Net Security / SecurityWeek

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…