Social learning, reimagined

Build social confidence through science & play.

Wavelength Learning is researching an adaptive, game-based system that learns when to support a learner—and when to preserve productive struggle—by studying the decisions of expert clinicians.

Explore our approach Partner with us

Proposed Phase I proof-of-mechanism research for learners in grades 6–8. No inference engine has yet been validated.

01Notice

02Choose

03Practice

04Reflect

Different paths.Meaningful growth.

Neurodiversity affirmingEvidence ledDesigned for real lifeBuilt with learners

Why the name Wavelength

There isn't one right wavelength.

Every learner sends signals in their own way. Our approach is about getting on their wavelength—paying attention to how they perceive, respond, and make meaning—then tuning support to the moment without treating one communication style as the standard.

Read the signalReason with careTune the support

Different patterns can still carry meaning.

Why Wavelength

Social learning should feel like life—not a worksheet.

Real conversations are dynamic. Context shifts, cues are subtle, and there is rarely one perfect answer. We are building a space where learners can explore those moments without pressure.

The goal is not to teach one “right” way to communicate. It is to expand each learner’s choices, confidence, and ability to advocate for what they need.

The proposed learning loop

Perceive. Believe. Act.

A research architecture designed to reason under uncertainty before deciding whether—and how—to scaffold.

Perceive

Represent privacy-conscious interaction traces such as choice trajectories, response timing, revisions, and cursor dwell.

Believe

Maintain a personalized, calibrated belief about a learner's likely bottleneck instead of assigning a fixed label.

Act

Choose whether to scaffold using an objective recovered from expert clinician decisions—not a hand-authored rulebook.

The experience

A world that responds to how you learn.

Learners move through age-respectful scenarios, make choices, notice what changes, and reflect on the outcome. The environment can adjust support and challenge while preserving agency.

Scenario-based social communication practice
Flexible pacing and multiple valid paths
Support for reflection without judgment
Progress insights for learners and trusted adults

See how we build

Scenario 04Practice mode

Want to join our project?

Choose what feels right to say.

No single “perfect” answer. Explore what fits the moment.

The scientific question

Can expert scaffolding become a recoverable, testable objective?

Our central hypothesis is that expert clinicians' moment-to-moment decisions reflect a latent objective that can be computationally recovered through inverse reinforcement learning—and tested against simpler rules.

Meet the team