Contradicted claimThe source states that cache is faster and closer to the processor.
From explanation to action
A single report connects the source, graph, diagnosis, and next step.
Instead of returning a score, Tessarion shows what the learner said, what the source supports, which concept relationship matters, and what to do next.
- Source evidence remains visible.
- Concept dependencies explain why a gap matters.
- The next action follows the diagnosed failure.
Example reportCache hierarchy
Teach-back diagnosis
GroundedMisconception detected
“Cache is slower because it is smaller.”
Covered correctlyThe learner separated cache from main memory.
Evidence2 source references support this decision.
Selected routeSocratic tutoringRepair speed, proximity, and memory hierarchy before the next teach-back.
Core capabilities
Explore the parts that make the learning loop work.
Select a capability to see its mechanism, boundary, and learner benefit.
Evidence grounding
Every diagnosis keeps source-chunk or concept evidence references.
The learner can inspect why a claim was accepted, questioned, or rejected.
01source chunk 04
02learner claim
03grounding validator
04evidence-linked finding
What the learner gains
Each technical boundary produces a clear learning advantage.
01
Inspectable evidence
Every diagnosis points to source chunks or concept evidence.
02
Visible prerequisites
Graph paths show which missing concept affects the current explanation.
03
Targeted recovery
Omissions, shallow answers, and misconceptions lead to different actions.
04
Grounded progress
Mastery changes after a new explanation, not after completing a chat.