LittlePlyResearch notebook

Selected default: CM-5, full-history move prediction. Browser play uses the verified CM-5 weights and matching full-history encoder. The Lichess service is active on Apple MPS (Mac GPU), with CPU fallback disabled and a separate iterative Rust 1/3/5/7-ply mate helper. Browser search remains three plies. Automated challenges obey the existing quotas and fall back to local Sunfish/Stockfish sampling when online play is unavailable.

Source map / CM-5 default

From full history to a legal move.

Retain the exact checkpoint, encoder, helper and source identity.

  1. Encode complete game history from the standard start; no cropping.
  2. Continue CM-4 to curated-data CM-5; inspect the shared model/training code.
  3. Run matching CM-5 browser controller and worker; manifest, weights and licenses.
  4. Inspect CM-8 random initialization, Sunfish evaluation and original/rendered hashes.

Why return to CM-5?

CM-5 is the selected default again. The additional state-conditioned CM-7 and fresh-random CM-8 experiments did not establish stronger play. The stopped direct CM-5/CM-7 run recorded 24 CM-5 wins, 17 CM-7 wins and 17 draws in 58 completed games, with one interruption and 41 unplayed. This supports an empirical operating decision, not a definitive ranking.

CM-8 training · CM-8: 6W / 2D / 92L against Sunfish · CM-7 versus CM-8: 69 / 10 / 21 · Stopped CM-5 versus CM-7 comparison.

Next investigate sequential causal move-history versus FEN-like board-state inputs under controlled supervision, curriculum, initialization, compute and helper settings. These runs confound those factors and prove no causal advantage of either representation. FEN describes current state; repetition history may be unknown.

Historical state-model source map

Source map / current CM-7

Historical source map

Follow the current implementation first. Historical pipelines remain below with their original source versions.

  1. Prepare examples and curriculum: ordinary games, book-supported moves and exact endgame alternatives; EXP-10 and EXP-11 record actual exposures.
  2. Encode current state: 120 prefix slots and three target tokens. Read the format.
  3. Train the initial curriculum, then the follow-up. CM-7 is the selected checkpoint, not the terminal update.
  4. Evaluate against Sunfish: protocol, outcomes and limitations, plus all games.
  5. Serve on Lichess or run in the browser. Their helper limits differ; the policy diagram separates model and search.

Check artifact availability and reproduction instructions before running code. Public snapshots redact host-specific paths and exclude credentials.

Historical full-history pipeline

The following map documents earlier model families. It remains useful for reproducing those experiments; use the current links above for CM-7.

  1. Data preparation and tokenizer: verified dataset-manifest code hashes; source-square, destination-square, promotion triples. Obtain Lichess source data separately.
  2. Training: complete-history causal objective. Read the recorded run configuration and results.
  3. Sunfish harness and model adapter: legal-move selection and separately labeled bounded forced-mate search. All 20 PGNs and replays.
  4. Historical GPU Lichess service: immutable release 6133648b2926-c205ec8f7f77, supervised locally with systemd. Casual standard-start chess from bullet to classical, including zero-increment games. Up to four simultaneous games when capacity allows; bullet games run alone. UltraBullet, correspondence and variants are unsupported. Availability depends on the local host remaining awake and running. Open the editable challenge dialog. Multi-game admission/controller, scheduler and inference owner are tied to the release manifest. This source subset excludes host configuration and launcher; it is not a complete deployment bundle. The CPU-service snapshot and initial bridge snapshot retain their historical limits. That historical snapshot retains its original checkpoint and evaluation; see the current release above.

Clean website build

npm ci
npm run check
npm run preview

Use Node 24.22.1. The lockfile pins dependencies. The build needs only the checked-in, sanitized notebook content and performs no training, match play, artifact retrieval or deployment. It emits static files to notebook/dist.

Reproducing model experiments

The exact recorded configuration, checkpoint digests, source hashes and PGNs are provided. Model-only exports and filtered raw/tokenized datasets are available. Resume checkpoints are excluded, and byte-exact pilot code is unavailable. Therefore this release is an inspectable experiment record, not yet a one-command reproduction bundle. Path-redacted source derivatives are labeled explicitly. Before a rerun, obtain licensed dependencies and data, restore local paths, and validate hashes against the recorded receipts. Do not interpret historical scripts with machine-specific launchers as portable run instructions.

Rust implementations

Preprocessing and token codec · Buildable codec source ZIP · Parallel mate search. Versioned source, tests, dependency locks and build guides; separate component licenses retained.