# IARPA The [[wiki/IARPA|Intelligence Advanced Research Projects Activity]] invests in high-risk, high-payoff research on difficult problems shared by the U.S. [[wiki/Intelligence Community|Intelligence Community]]. It sits within the [[wiki/Office of the Director of National Intelligence|Office of the Director of National Intelligence (ODNI)]]. A typical program takes a difficult intelligence question and makes it testable. Can an analyst assess a prediction before the answer is known? Can an image reveal what a site is used for? Can a sensor distinguish an authentic identity sample from a presentation attack? The public program record lets a reader follow those questions through performers, benchmarks and reported results. Transition to operational work has its own evidence. [About IARPA](https://www.iarpa.gov/who-we-are/about-us) · [IARPA research programs](https://www.iarpa.gov/research-programs). [[collections/DARPA|DARPA]] is the companion defense research collection. IARPA's through-line is the quality of evidence and judgment across the intelligence system. Follow it into [[collections/Terrorism, Counterterrorism, and the Intelligence Environment|counterterrorism]], [[wiki/Austin Defense Technology|Austin defense technology]], [[collections/Neurotech|neurotechnology]] and [[collections/Gamification|gamification and forecasting]]. [[#Program and challenge index|Browse the program and challenge index]]. ## Directional routes through the corpus ### Forecasting and analyst judgment **Question → probability estimate → independent scoring → calibration and aggregation → decision support.** [[wiki/Aggregative Contingent Estimation|ACE]] studied how to improve and combine forecasts about real-world events. The [[wiki/Good Judgment Project|Good Judgment Project]] was a performer team within that research, with a public forecast dataset released through IARPA. Sirius used serious games to study mitigation of cognitive biases. REASON asked whether tools could help analysts evaluate the evidence and reasoning in draft reports. These programs address distinct points in the same workflow. [IARPA ACE](https://www.iarpa.gov/research-programs/ace) · [IARPA Good Judgment data release](https://www.iarpa.gov/newsroom/article/iarpa-announces-publication-of-data-from-the-good-judgment-project). - [[articles/gamification/Lex Sportiva and Winning Without Bloodshed|Lex Sportiva and Winning Without Bloodshed]] connects ACE and the Good Judgment Project to the tournament as a way to measure performance under agreed rules. This is the corpus entry into [[wiki/Forecasting Tournament|forecasting tournaments]] and [[wiki/Superforecaster|superforecasting]]. - [[articles/Peak Person and the Predicaments of Prediction|Peak Person and the Predicaments of Prediction]] supplies the adjacent question of how to use a prediction after it is made. It belongs downstream of the forecasting research and its scores. - [[collections/DARPA#Information Awareness Office|The DARPA Information Awareness Office route]] gives the historical comparison with FutureMAP's prediction-market research. The documented relationships are within each program. ### Evidence, interpretation and counterterrorism **Observed signal → integrity or provenance check → interpreted evidence → uncertainty-aware inference.** [[wiki/IARPA Odin|Odin]] asked whether an identity sensor could tell a valid presentation from an attack. [[wiki/Functional Map of the World|fMoW]] asked teams to classify the functions of sites in satellite imagery. [[wiki/Deep Intermodal Video Analytics|DIVA]] studied video understanding under benchmark conditions. AGILE addressed computing for large, sparse and changing data, with graph analytics among the relevant workloads. The [[wiki/Counterterrorism Predictive Graph|counterterrorism predictive graph]] is the corpus model for how evidence at those different stages can inform a changing estimate of a network. - [[collections/Terrorism, Counterterrorism, and the Intelligence Environment|Terrorism, Counterterrorism, and the Intelligence Environment]] supplies the operational and institutional graph. Read [[wiki/Counterterrorism Collection - Predictive Graphs|Predictive Graphs]] for data association, missing observations and probabilities. The research records here describe the component problems and benchmarks. - [[articles/Cognitive-Cyber Warfare Measures and Countermeasures|Cognitive-Cyber Warfare Measures and Countermeasures]] places ReSCIND in the defensive research lineage and uses HIATUS to discuss the evidence and uncertainty in authorship attribution. [[collections/DARPA#From observation to evidence and judgment|DARPA's media and event reasoning route]] is its cross-agency companion. - [[articles/Person of Interest and The Machine#The Named Cognitive Core|Is Person of Interest's “The Machine” Real?]] draws on public research to explore a fictional architecture that relates perception and inference. The article is a conceptual comparison. The benchmarks and performer records below establish the individual programs. ### Austin, biometrics and imagery **Biometric evidence → presentation-attack detection → evaluator → performer and region.** IARPA announced [[wiki/IARPA Odin|Odin]] on **October 19, 2017** to research detection of attacks on face, iris and fingerprint authentication. Thor and Loki were solicitations under Odin, and IARPA names [[wiki/HID Global|HID Global]] as one of four prime performers. The HID link gives the Austin research map a named company and a named award relationship. The corpus uses [[wiki/Biometric Presentation Attack Detection|presentation-attack detection]] to explain the technical question. [Odin program record](https://www.iarpa.gov/research-programs/odin) · [IARPA launch announcement](https://www.iarpa.gov/newsroom/article/iarpa-launches-odin-program-to-harden-biometric-technology-against-attacks). - **Phonetic and semantic disambiguation.** The [[wiki/ODEN, ODIN, and ODNI|Oden, Odin, ODIN and ODNI route]] distinguishes the UT institute, the IARPA program, the Army task force, and the office that houses IARPA. - **MOSAIC.** [[wiki/IARPA MOSAIC|IARPA MOSAIC]] studied multimodal assessment of cognitive and psychological characteristics in context. [[wiki/Dallas MOSAIC|Dallas MOSAIC]] belongs to regional information sharing. - [[research/The Austin Executable Loop|The Austin Executable Loop]] places biometrics, state estimation, sensing and regional institutions in one evidence map. [[wiki/Austin Defense Technology#Institutional and product distinctions|Austin Defense Technology]] gives the reader the shortest route between the Odin, ODIN and Oden records. - [[articles/How Austin Became a Different Kind of Defense Tech Epicenter#The Executable Loop|How Austin Became a Different Kind of Defense Tech Epicenter]] supplies the regional engineering context for this comparison. For the actual IARPA performer evidence, follow Odin and HID Global here. [[collections/DARPA#Austin, semiconductors and manufacturing|DARPA's TIE award]] is a separate Austin funding edge. - [[articles/The Coming Market for Civilian Spycraft|The Coming Market for Civilian Spycraft]] describes the ODNI-connected data marketplace where government looks for commercial inputs. It is a related channel for acquiring data and capability, with research awards and performer relationships documented in the IARPA records below. ### Biological reference maps and machine intelligence **Record function → reconstruct structure → register both → ask what the circuit computes.** [[wiki/MICrONS|MICrONS]] paired measurements of activity in mouse visual cortex with structural reconstruction of the same tissue. The public [[wiki/MICrONS Explorer|MICrONS Explorer]] and the published connectomics dataset let readers follow the scientific output beyond the funding announcement. [[collections/Neurotech|Neurotech]] tracks that work alongside [[collections/DARPA#Neural access, rehabilitation and brain maps|DARPA's neural interface programs]]. [IARPA MICrONS](https://www.iarpa.gov/research-programs/microns). - [[articles/Machine Intelligence from Cortical Networks and the Allen Institute#Machine Intelligence from Cortical Networks|Machine Intelligence from Cortical Networks and the Allen Institute]] is the focused reading on the consortium's scientific output and why a paired structure–function map matters for machine intelligence. - [[articles/The Organic-Synthetic Brain Atlas#III.4 — MICrONS: Structure-Function Fusion in Mouse Visual Cortex|The Organic-Synthetic Brain Atlas]] places MICrONS between structural connectomics and methods that connect wiring to measured dynamics. - [[articles/2026 Annual Report on Brain-Computer Interfaces#MICrONS: Function-Structure Registration at Mammalian Scale|2026 Annual Report on Brain-Computer Interfaces]] uses MICrONS as a reference map in the larger BCI ecosystem. The research output supplies a model of circuitry. The article supplies the bridge to interface engineering. ### Language, data and trusted computation **Raw speech or records → representation → discovery across datasets → usable evidence.** Babel studied speech recognition in languages with scarce training resources. KDD addressed discovery across heterogeneous information. HIATUS investigated explainable authorship attribution and authorship privacy. AGILE addressed computational architectures that can handle large-scale analytic workloads. HECTOR studied techniques for using advanced cryptography in distributed applications. These are language, evidence and computing programs with distinct deliverables. Follow [[wiki/Graph-Based Intelligence Analysis|Graph-Based Intelligence Analysis]] for the corpus's analytic layer, and [[wiki/IC ITE|IC ITE]] for the separate shared-services and enterprise infrastructure layer. [IARPA program history](https://www.iarpa.gov/who-we-are/history/our-programs). ## Program and challenge index The entries below route to the existing vault note where one exists and otherwise to IARPA's public record. The subjects describe the research questions. IARPA's catalog gives its own office and status assignments. The Functional Map of the World was announced as a challenge. The Good Judgment Project was an ACE performer. **“Social Threat Observatories”** remains a phrase from prior discussion with an unresolved program attribution. Catalog reviewed November 20, 2026. [IARPA program directory](https://www.iarpa.gov/research-programs) · [IARPA past programs](https://www.iarpa.gov/research-programs?show_current_past=past&show_office=all). ### Forecasting, judgment and analyst reasoning - [[wiki/Aggregative Contingent Estimation|Aggregative Contingent Estimation (ACE)]] — Methods for improving and aggregating probability judgments. [IARPA record](https://www.iarpa.gov/research-programs/ace). - **Crowdsourcing Evidence, Argumentation, Thinking and Evaluation (CREATE)** — Tools for assessing hypotheses and reasoning in teams. [IARPA record](https://www.iarpa.gov/research-programs/create). - [Mercury](https://www.iarpa.gov/research-programs/mercury) - **Open Source Indicators (OSI)** — Early indicators extracted from open sources. [IARPA record](https://www.iarpa.gov/research-programs/osi). - **Sirius** — Serious games and evaluation of training to mitigate cognitive bias. [IARPA record](https://www.iarpa.gov/research-programs/sirius). - **Rapid Explanation, Analysis and Sourcing Online (REASON)** — Evidence and reasoning support for analysts drafting intelligence reports. [IARPA record](https://www.iarpa.gov/research-programs/reason). - [Strengthening Human Adaptive Reasoning and Problem-solving (SHARP)](https://www.iarpa.gov/research-programs/sharp) - [Forecasting Counterfactuals in Uncontrolled Settings (FOCUS)](https://www.iarpa.gov/research-programs/focus) - [Forecasting Science & Technology (ForeST)](https://www.iarpa.gov/research-programs/forest) - [Hybrid Forecasting Competition (HFC)](https://www.iarpa.gov/research-programs/hfc) - [[wiki/Good Judgment Project|Good Judgment Project]] — performer within ACE. Its forecast dataset supports independent study of training, aggregation and scoring. [IARPA data release](https://www.iarpa.gov/newsroom/article/iarpa-announces-publication-of-data-from-the-good-judgment-project). ### Language, text and knowledge discovery - **Babel** — Speech recognition in languages with limited training data. [IARPA record](https://www.iarpa.gov/research-programs/babel). - **Human Interpretable Attribution of Text using Underlying Structure (HIATUS)**. Connects interpretable attribution and authorship privacy to the evidence used in assessing information campaigns. [[articles/Cognitive-Cyber Warfare Measures and Countermeasures|Cognitive-Cyber Warfare Measures and Countermeasures]]. [IARPA record](https://www.iarpa.gov/research-programs/hiatus). - **Knowledge Discovery and Dissemination (KDD)** — Discovery and analysis across disparate datasets. [IARPA record](https://www.iarpa.gov/research-programs/kdd). - [Machine Translation for English Retrieval of Information in Any Language (MATERIAL)](https://www.iarpa.gov/research-programs/material) - [Metaphor](https://www.iarpa.gov/research-programs/metaphor) - [Foresight and Understanding from Scientific Exposition (FUSE)](https://www.iarpa.gov/research-programs/fuse) - [Socio-cultural Content in Language (SCIL)](https://www.iarpa.gov/research-programs/scil) - [Tools for Recognizing Useful Signals of Trustworthiness (TRUST)](https://www.iarpa.gov/research-programs/trust) ### Biometric assurance and human-state research - [[wiki/IARPA Odin|Odin]] — Biometric presentation-attack detection in face, iris and fingerprint systems. [IARPA record](https://www.iarpa.gov/research-programs/odin). - [[wiki/IARPA MOSAIC|Multimodal Objective Sensing to Assess Individuals with Context (MOSAIC)]] — Multimodal measurement of individual characteristics in context. [IARPA record](https://www.iarpa.gov/research-programs/mosaic). - [Janus](https://www.iarpa.gov/research-programs/janus) - [Biometric Recognition and Identification at Altitude and Range (BRIAR)](https://www.iarpa.gov/research-programs/briar) - [Biometrics Exploitation Science & Technology (BEST)](https://www.iarpa.gov/research-programs/best) - [Targeted Evaluation of Ionizing Radiation Exposure (TEI-REX)](https://www.iarpa.gov/research-programs/tei-rex) ### Geospatial, event and video understanding - [[wiki/Deep Intermodal Video Analytics|Deep Intermodal Video Analytics (DIVA)]] — The research program and its connection to video benchmarks and activity detection. [IARPA record](https://www.iarpa.gov/research-programs/diva). - [Creation of Operationally Realistic 3D Environment (CORE3D)](https://www.iarpa.gov/research-programs/core3d) - **Space-based Machine Automated Recognition Technique (SMART)** — Broad-area remote-sensing analysis of change over time. [IARPA record](https://www.iarpa.gov/research-programs/smart). - [Finder](https://www.iarpa.gov/research-programs/finder) - [Aladdin Video](https://www.iarpa.gov/research-programs/aladdin-video) - [Walk-through Rendering from Images of Varying Altitude (WRIVA)](https://www.iarpa.gov/research-programs/wriva) - [Video Linking and Intelligence from Non-Collaborative Sensors (Video LINCS)](https://www.iarpa.gov/research-programs/video-lincs) - [Location Using Sound (LocUS)](https://www.iarpa.gov/research-programs/locus) - [Commercial Observation for Spatio-temporal Monitoring for Indications of Change (COSMIC)](https://www.iarpa.gov/research-programs/cosmic) - [Space Debris Identification and Tracking (SINTRA)](https://www.iarpa.gov/research-programs/sintra) - [Movement Observation and Video-based Evaluation System (MOVES)](https://www.iarpa.gov/research-programs/moves) - [[wiki/Functional Map of the World|Functional Map of the World Challenge (fMoW)]] — Public challenge on classifying site functions from satellite imagery. [IARPA record](https://www.iarpa.gov/newsroom/article/iarpa-announces-functional-map-of-the-world-challenge-to-develop-automated-analysis-of-satellite-imagery). ### Computing, cryptography and graph infrastructure - **Advanced Graphic Intelligence Logical Computing Environment (AGILE)** — Energy-efficient architectures for large, sparse and changing analytic workloads. [IARPA record](https://www.iarpa.gov/research-programs/agile-2). - **Homomorphic Encryption Computing Techniques with Overhead Reduction (HECTOR)** — Advanced cryptography and more practical secure distributed computing. [IARPA record](https://www.iarpa.gov/research-programs/hector). - [Security and Privacy Assurance Research (SPAR)](https://www.iarpa.gov/research-programs/spar) - [Coherent Superconducting Qubits (CSQ)](https://www.iarpa.gov/research-programs/csq) - [Quantum Computer Science (QCS)](https://www.iarpa.gov/research-programs/qcs) - [Multi-Qubit Coherent Operations (MQCO)](https://www.iarpa.gov/research-programs/mqco) - [Quantum Enhanced Optimization (QEO)](https://www.iarpa.gov/research-programs/qeo) - [Entangled Logical Qubits (ELQ)](https://www.iarpa.gov/research-programs/elq) - [Logical Qubits (LogiQ)](https://www.iarpa.gov/research-programs/logiq) - [SuperCables](https://www.iarpa.gov/research-programs/supercables) - [SuperTools](https://www.iarpa.gov/research-programs/supertools) - [MicroElectronics for Artificial Intelligence (MicroE4AI)](https://www.iarpa.gov/research-programs/microe4ai) - [Artificial Reasoning for Circuit Automation and Design Engineering (ARCADE)](https://www.iarpa.gov/research-programs/arcade) - [Trusted Integrated Chips (TIC)](https://www.iarpa.gov/research-programs/tic) - [Cryogenic Computing Complexity (C3)](https://www.iarpa.gov/research-programs/c3) - [Circuit Analysis Tools (CAT)](https://www.iarpa.gov/research-programs/cat) - [Rapid Analysis of Various Emerging Nanoelectronics (RAVEN)](https://www.iarpa.gov/research-programs/raven) - [DECIPHER](https://www.iarpa.gov/research-programs/decipher) ### AI assurance and defensive systems - [Cyber-attack Automated Unconventional Sensor Environment (CAUSE)](https://www.iarpa.gov/research-programs/cause) - [Securing our Underlying Resources in Cyber Environments (SoURCE CODE)](https://www.iarpa.gov/research-programs/source-code) - **BENGAL** — Large language model evaluation and threat characterization. [IARPA record](https://www.iarpa.gov/research-programs/bengal). - **Reimagining Security with Cyberpsychology-Informed Network Defenses (ReSCIND)**. Reads ReSCIND as a defensive research program on attacker decision-making. [[articles/Cognitive-Cyber Warfare Measures and Countermeasures|Cognitive-Cyber Warfare Measures and Countermeasures]]. [IARPA record](https://www.iarpa.gov/research-programs/rescind). - [Virtuous User Environment (VirtUE)](https://www.iarpa.gov/research-programs/virtue) - [Scientific advances to Continuous Insider Threat Evaluation (SCITE)](https://www.iarpa.gov/research-programs/scite) - [Securely Taking On New Executable Software of Uncertain Provenance (STONESOUP)](https://www.iarpa.gov/research-programs/stonesoup) ### Physical sensing, signals and research neuroscience - [Little Horned Owl (LHO)](https://www.iarpa.gov/research-programs/lho) - [Robust Energy Sources for Intelligence Logistics In Extreme Novel and Challenging Environments (RESILIENCE)](https://www.iarpa.gov/research-programs/resilience) - [Superior Options for Long-life Solar Technologies with Impressive Conversion Efficiencies (SOLSTICE)](https://www.iarpa.gov/research-programs/solstice) - [Pursuing Intelligent Complex Aerosols for Rapid Detection (PICARD)](https://www.iarpa.gov/research-programs/picard) - [Signal Location in Complex Environments (SLICE)](https://www.iarpa.gov/research-programs/slice) - [[wiki/MICrONS|Machine Intelligence from Cortical Networks (MICrONS)]] — Paired functional and structural maps from mouse visual cortex. [IARPA record](https://www.iarpa.gov/research-programs/microns). - [Knowledge Representation in Neural Systems (KRNS)](https://www.iarpa.gov/research-programs/krns) - [Gaseous Radioisotope Analysis in Situ Laboratory (GRAIL)](https://www.iarpa.gov/research-programs/grail) - [Effective Quantitative Antenna Limits for Performance (EQuAL-P)](https://www.iarpa.gov/research-programs/equal-p) - [Standoff ILluminator for Measuring Absorbance and Reflectance Infrared Light Signatures (SILMARILS)](https://www.iarpa.gov/research-programs/silmarils) - [Molecular Analyzer for Efficient Gas-phase Low-power INterrogation (MAEGLIN)](https://www.iarpa.gov/research-programs/maeglin) - [High Frequency Geolocation (HFGeo)](https://www.iarpa.gov/research-programs/hfgeo) - [Securing Compartmented Information with Smart Radio Systems (SCISRS)](https://www.iarpa.gov/research-programs/scisrs) - [Ithildin](https://www.iarpa.gov/research-programs/ithildin) - [Amon-Hen](https://www.iarpa.gov/research-programs/amon-hen) - [PROTEIN-BASED HUMAN IDENTIFICATION METHODS (Proteos)](https://www.iarpa.gov/research-programs/proteos) - [Bio-Intelligence Chips (BIC)](https://www.iarpa.gov/research-programs/bic) - [Smart Electrically Powered and Networked Textile Systems (SMART ePANTS)](https://www.iarpa.gov/research-programs/smart-epants) ## Relationships - **institution:** [[wiki/IARPA|IARPA]] and its [[wiki/Intelligence Advanced Research Projects Activity|full-name entry]]. - **parent and research customer community:** [[wiki/Office of the Director of National Intelligence|ODNI]] and the [[wiki/Intelligence Community|Intelligence Community]]. - **defense research counterpart:** [[collections/DARPA|DARPA]]. - **evidence and inference in use:** [[collections/Terrorism, Counterterrorism, and the Intelligence Environment|Terrorism, Counterterrorism, and the Intelligence Environment]]. - **regional performer evidence and nomenclature:** [[wiki/Austin Defense Technology|Austin Defense Technology]] and [[wiki/ODEN, ODIN, and ODNI|Oden, Odin, ODIN and ODNI]]. - **structure and function in neural research:** [[collections/Neurotech|Neurotech]]. - **forecast scoring and calibration:** [[collections/Gamification|Gamification]] and [[wiki/Good Judgment Project|Good Judgment Project]].