Preserved version of Cognivia as it appeared on September 3, 2026. Actions are disabled here. Research Platform Live site

Cognivia · Cognitive diagnostics for learning

Every mindforgets differently.

A learning-diagnostics system, built on cognitive science. It measures what a learner's memory is doing behind an answer, and hands back a reading any teaching tool can act on.

What Cognivia measures

Cognivia measures what a learner's memory is doing behind an answer: how firmly they hold an idea, and how fast it is slipping, read live while they work.

Right now it is an independent research project, and soon a public-benefit startup, built for students, teachers, and researchers in rural government schools. Whatever it becomes, its job stays the same: to measure. It leaves scheduling, grading, and lesson content to the tools a class already uses.

Observed37 retrieval events on this concept, over 41 days.
Inferred · with certaintyHigher chance of rapid forgetting, retention 0.42 ± 0.09.
Recommended · then verifiedRetrieval → delayed feedback → transfer check. Re-measured after.

The rule that keeps it honest: every estimate ships with its uncertainty and the evidence behind it, so a reader can weigh it instead of taking it on trust.

See the intelligence layer →

Why memory fades

Memory decays along a precise curve.

It does not fade randomly. It fades along a steep, regular curve that scientists have mapped for well over a century. Almost no classroom has changed because of it. Cognivia treats learning as something you can measure: a tool that reads how knowledge forms and fades in each learner, then adapts to what it finds.

Retention over 14 days · drag the review points
Ebbinghaus 1885 · live model
Retention · day 14
Drag the ● review points ↔
No review With your reviews

Where memory becomes measurable

Reading the signal

See exactly how forgetting happens.

Every answer, how long it took, and how sure the student felt feeds a forgetting model built just for them. Within one session, their own curve of forgetting starts to show, the signal no exam ever captures.

At the moment it happens

Read what kind of wrong it is.

From the timing and the confidence behind an answer, the confirm-or-clear diagnostic works out which kind of wrong it is, then routes it to the fix that kind needs.

Acting on it

Let the diagnosis pick the next move.

Show a misconception a counter-example that breaks it. Send a fading memory back for retrieval before it goes. A missing prerequisite means stepping back a level first, then returning. Each move is checked at the next review, so a fix that did not land gets caught instead of assumed.

A system, not a feature

Cognivia keeps measuring after the answer.

The reading keeps updating. It refreshes every session, holds a separate reading for every topic, and follows the learner from one class to the next.

So a teacher who meets a student in a new year inherits that state instead of a blank slate, and nothing quietly resets at term's end. In the Cognitive Passport it becomes one page per learner: what is holding, what is slipping, and how fast.

Four kinds of wrong

Four ways to be wrong. Each one needs a different fix.

How long a wrong answer took, and how sure the student was, place it in one of four kinds. When two trained people label the same errors and agree on the kind, the sorting counts as a real measurement, not a guess. This reading is the part no competitor has.

01

Plain forgetting

Looks like: held it before, now slower and less sure.

Fix: retrieval, timed to how fast this idea is slipping.

02

A missing prerequisite

Looks like: fails on something it depends on but never held.

Fix: step back to the prerequisite, then return.

03

A half-formed grasp

Looks like: right sometimes, hesitant, breaks when the wording changes.

Fix: worked examples to consolidate before testing.

04

A confident misconception

Looks like: fast, certain, and wrong, the same wrong answer each time.

Fix: a contrasting counter-example, never more drilling.

Cognivia · The diagnostic engine, live

Watch it read five
answers, and diagnose.

Answer five questions and watch the engine work: it sorts each error by type and builds your Learning Genome right in the browser, updating as you go.

Question 1 of 5

Two students read the same page for the same hour
and forget it on entirely different days.

What replaces the transcript

A record of what a learner will still remember next month.

The Old Document

"What did you score?"

It is a number and a rank, an average of averages, built to sort students rather than understand them. It squeezes everything you are as a learner into one figure, so you can be lined up against the next student.

The Cognitive Passport

"How does your memory actually work?"

A profile built from how you work through problems, session after session. It grows with you and keeps the thing a report card throws away: how you actually reached the answer, not just whether it was right.

Cognitive Passport

Biology · Class 12

Updated this session

Learning Genome · this subject

λForgetting rateslow
RSRetrieval strength0.81
CEConsolidation efficiency0.74
FSIFatigue sets in~7 min

Scheduled review

PhotosynthesisHeld · retrieval strong
in 6 days
Chloroplast structureFading · due for retrieval
in 2 days
Chlorophyll & lightMisconception flagged
tomorrow
Calvin cycleNew · first review
in 1 day

One page per learner: what is holding, what is slipping, how fast, and why. This is a representative view, drawn from sample data.

What changes for teaching

The fix changes with the kind of wrong.

Because Cognivia names the kind of error, the same wrong answer from two students can earn two different responses. Practice stops being one-size-fits-all, and an hour of review lands where it actually helps.

4Failure types Cognivia tells apart, so the right fix follows.

Alongside your stack

It works alongside the tools you already run.

Cognivia takes one thing a learner did and returns why it happened, what is likely to happen next, and what to do about it, through one simple API. A school keeps its LMS, a tutor keeps their app, and Cognivia adds the piece they were missing: a real reading of memory.

10Variables in the Learning Genome, built from a learner's own events.

For institutions, classrooms & teams

See what the whole class actually remembers.

Add learners into groups with role-based access, owner, teacher, learner. You see who is struggling and, more useful, why, so two students who scored the same can each get the fix they actually need. You catch the one quietly slipping while there is still time to change how you teach.

Grade 10 · Biology
28 learners · updated this session
OwnerTeacherLearner

Learners · tap one

Integrations

The memory state behind every answer, handed back to the tools you already run.

Your LMS, gradebook, and tutor already deliver material and record what a learner did. None of them measure what the learner's memory is actually doing. Cognivia measures that and hands the reading back through one clean API and standard LTI, so it shows up wherever you already work.

The memory-state API

Every learner's live cognitive state, one query away, ready to read wherever your teaching already lives. The diagnosis and reports come back in whatever language your learners use.

// a learner's live memory state
GET /v1/learners/:id/memory-state

{ "concept": "Photosynthesis",
"retrievability_now": 0.91, "fsrs_stability": 7.6,
"accuracy_ci95": [0.21, 0.94],
"next_review_at": "2026-08-30" }

Plugs into what you already use

Standard LTI 1.3 for single sign-on and roster sync, plus webhooks that fire the moment a learner's state changes, so the right nudge reaches the right place.

LTI 1.3REST APIWebhooksCSV exportRoster syncSSOAny language

Research

Most learning tools open with their best number. Here are ours, before the trial has run.

Cognivia is tested in a pre-registered, three-arm randomised trial. The design, the analysis plan, and the data are public before the results are known. You will not find a "4× faster" or a "90% retention" on this page, because those are outcomes, and the outcome is not in yet. When it is, it goes up whether or not it helps our case.

0enrollment target
0randomised arms · pre-registered
2-wkprimary endpoint
OSFpre-registered, public

Every formula and threshold Cognivia runs on is written down and public, so you can read the method instead of trusting a black box. The science is named: Ebbinghaus (1885) on the shape of forgetting, and Wixted & Carpenter (2007) and Bjork & Bjork (1992) on how memory strengthens and fades.

OSF pre-registration → CFI protocol preprint (EdArXiv) → Retention study preprint → SSRN →

Registered as a Stage 1 pre-registered protocol; the analysis plan was locked before any data was collected. Every link above opens a real, public document you can read right now.

A student is not a 78%.

A student is a particular forgetting rate on photosynthesis, a stable retrieval strength on chloroplasts, a fatigue threshold at seven minutes, and an active misconception about chlorophyll that needs to be flagged before tomorrow.

See what your memory
is actually doing.

In one session, Cognivia reads why you miss what you miss, and what to do about it. Free for students, built in Kanpur, tested in the open.