LittlePlyModel + harness

Research / evidence before conclusions

What changed. What held up.

Model training and harness design are separate lines of work. Each result belongs to its own checkpoint, configuration and protocol.

Act Two / paused indefinitely

PVM-01: learning to use a chess VM.

Act Two is indefinitely paused. PVM-01 remains an experimental checkpoint, not the live player. The Act One CM-14 agent continues on Lichess.

Read the early results and unrun GRPO plan →
Recorded / CPU

Same model, a better harness.

The CM-14 development record moves from 72% in a 100-game mate-only cohort to 99% in a later 50-game full-system cohort. Protocols differ; this is not a controlled search-only gain.

Sunfish results and dated Lichess ratings →
Unsuccessful replacement

CM-16’s multi-day experiment.

Auxiliary objectives, adaptation and continuation did not establish a stronger replacement for CM-14. The compute-budget explanation remains a working hypothesis.

Outcome and limitations →
Earlier experiment

Looking farther ahead in training.

CM-15 tested a future-move target. Its completed initial evaluation did not establish a strength gain over the matched reference.

CM-15: methods and games →
In development / no strength claim

Make inference-time work count.

Async GPU search, opponent-turn work and cross-game batching are the next harness direction. Historical CPU results do not validate this implementation.

The design and its tests →

Read at your preferred depth

New to the terminology?

How to read an evaluation explains score, win rate, ablations and rating caveats.

Want to reproduce a result?

Choose a reproduction target before downloading. Published snapshots retain their dates and known gaps.