RAM (Restoring Active Memory) Replay | Apply Online

FAQs: DARPA RAM (Restoring Active Memory) Replay Opportunity (BAA 15-34)

What is the DARPA RAM Replay opportunity (BAA 15-34)?

It is a DARPA Biological Technologies Office (BTO) research and development opportunity focused on human neuroscience, specifically the role of "memory replay" in consolidating episodic memories and newly learned skills, and how replay-driven memories are later recalled and used during real task performance.

What core scientific question is this program trying to answer?

The program centers on how memory replay supports (1) consolidation of episodic memories and newly learned skills and (2) later recall and use of those memories during real task performance.

What does DARPA mean by moving beyond descriptive studies of memory?

Based on the opportunity description, DARPA is looking for teams that do more than observe or describe replay. The emphasis is on building quantitative, testable models of replay that produce measurable parameters and can be validated against data and outcomes.

What kinds of outputs is DARPA looking for from proposed models?

The models should yield actionable metrics tied to replay, such as measurable parameters for when replay occurs, indicators of replay quality, how replay relates to later recall accuracy, and how replay differs across individuals or task conditions.

What are examples of replay-related parameters or variables that proposals might quantify?

The announcement highlights examples such as: when replay occurs; neural or behavioral signatures indicating replay quality; relationships between replay and later recall accuracy; and variation in replay across individuals or task conditions.

What does "actionable metrics" mean in the context of this program?

It refers to metrics that can be estimated from measurable data streams and used to assess replay quality in a way that can inform decisions or interventions aimed at improving later recall and real task performance.

What kinds of data streams does the opportunity suggest could be used to estimate replay metrics?

The description lists plausible experimental and applied data streams such as neural recordings, physiological signals, task performance data, or other quantifiable indicators.

How should proposals connect replay metrics to outcomes?

Proposals are expected to link replay-related metrics to downstream outcomes like subsequent memory retrieval and performance, rather than treating replay measures as standalone descriptive features.

Is the program only about understanding replay, or also about improving memory?

It is explicitly framed as both: understanding replay and translating that understanding into methods that assess replay quality and enable, strengthen, or improve memory recall in operationally relevant settings.

Does the opportunity expect interventions to be part of the proposed work?

Yes, the opportunity emphasizes using the model as a foundation for interventions. Proposals should describe how modeling would drive strategies to evaluate whether replay is occurring effectively and how to improve recall and task performance that depend on replay.

Are specific intervention modalities required or listed?

No specific intervention modalities are listed in the provided description. The framing suggests interest in interventions that can be empirically tested, compared against baselines, and evaluated for effects on recall and performance in tasks resembling real-world operational demands.

What does the opportunity imply by "closed-loop or decision-guided strategies"?

It implies approaches where replay metrics are used to guide evaluation and improvement strategies, potentially in a feedback-driven (closed-loop) manner or through decision rules informed by the model and measured replay quality.

What types of tasks does DARPA want used for evaluation?

The description suggests that evaluation should focus on tasks resembling real-world operational demands rather than purely artificial lab tasks.

How many awards did DARPA expect to make?

The expected number of awards was around 10.

What is the funding category for this opportunity?

It was a discretionary federal R&D opportunity in the Science and Technology/Research and Development category.

What award instruments might DARPA use for this program?

DARPA indicated multiple possible award instruments: cooperative agreements, other transactions, and procurement contracts.

What does it mean that multiple award instruments are possible?

In practical terms, it signals flexibility in how projects may be managed and transitioned depending on the technical approach and maturity, since the opportunity allows for different instrument types.

Was cost sharing or matching required?

No. The description states there was no cost sharing or matching requirement.

Who was eligible to apply?

Eligibility was listed as unrestricted, meaning any qualified entity type could apply, subject to any additional clarifications in the full announcement.

What is the CFDA number associated with this opportunity?

The CFDA number listed was 12.910 (Research and Technology Development).

When was the opportunity posted?

It was posted in late April 2015.

What was the application deadline?

The closing deadline was June 16, 2015 at 4:00 pm.

When was the archive date?

The archive date was July 16, 2015.

Who was the point of contact for access issues related to the full announcement package?

The listed point of contact was Dr. Justin Sanchez, Program Manager, DARPA BTO, through the BAA coordinator.

What office at DARPA ran this program?

The program was run through DARPA's Biological Technologies Office (BTO).

What is the program trying to achieve in operational terms?

Operationally, the program aims to develop methods to assess replay quality and translate those assessments into interventions that improve later memory recall and real task performance in settings relevant to operational demands.

What is a key requirement emphasized in the announcement for proposers?

A central requirement is proposing novel approaches to modeling replay that produce actionable metrics, including ways to estimate and validate replay-related variables from plausible data streams and link them to recall and performance outcomes.