RESULTS.md — append-only

Every number the project reports lives here, in order. Never edit a past row; add a new one. Scores are the mean binary task score on the named split (tasks/README.md). Cost is input+output tokens through the Anchor's proxy.

Environment

itemvalue
hostDebian 13, Xeon E5-1620 v2 (8 threads, no GPU), 31 GiB RAM
sandbox imagepython:3.12-slim @ sha256:2fe5997d…ce79, pytest 9.1.1
sandboxrootless Podman 5.4.2, --network=none --memory=1g --cpus=2
modelornith-1.5:9b (qwen35 arch, Q4_K_M, 8.95 B) on local Ollama 0.33.2
model settingstemperature 0, seed 0, thinking off, context 16384, one request at a time

Runtime choice (2026-09-09)

Same GGUF, same CPU, measured with llama-bench (llama.cpp build 434ddbb, -march=native) and Ollama's /api/generate timings:

runtimethreadsprompt eval tok/sgeneration tok/s
Ollama 0.33.247.44.07
llama.cpp native47.26 ± 0.034.07 ± 0.01
llama.cpp native88.00 ± 0.113.60 ± 0.13

No measurable difference: Ollama runs the same AVX kernels (this CPU has no AVX2/FMA). Ollama stays as the runtime; its OpenAI-compatible endpoint supports tool calls and reasoning_effort: "none". Prompt processing is the dominant cost, so the harness keeps message history append-only to reuse the KV cache between steps.

Phase 1 — task suite

(filled in when the suite is frozen)

Phase 2 — Anchor acceptance (2026-09-09)

criterionevidence
zero LLM calls, no credentialsTestNoModelCalls: no endpoint literals, no provider SDK, net/http only in proxy.go
size987 non-blank non-comment LOC (TestAnchorSize, ceiling 1000; target 500 not yet met)
sandbox cannot read anchor/ or holdout/TestSandboxIsolation: real container, walks /, finds no anchor, holdout, reference, test_hidden.py; no network; writes outside /work fail
rollback reproduces the scoreTestRollbackRestoresSnapshotAndScore: cache hit, identical train/holdout, working copy hash equals the snapshot
acceptance ruleTestApplyRejections (8 rejection reasons) and TestApplyAccept
reconstructible ledgerTestLedgerReconstructsGeneration: parent snapshot + stored diff hashes to the generation's snapshot

go test -race ./... green on this host with rootless Podman 5.4.2.

Phase 3 — feasibility run (2026-09-09, before the suite was frozen)

One real end-to-end run through the Anchor (sandbox, proxy, Ollama ornith-1.5:9b) on the exemplar task 001_word_frequency:

scorehidden testsstepswalltokens in/out
1.019/194274 s6596 / 571

The model read the file, ran the visible tests, rewrote the file with a sorted-key fix and re-ran the tests, then stopped. A first attempt with the task's original 300 s timeout was killed mid-step-2; task timeouts are 1800 s from here on. Not a baseline number: one task, train split, cache warm.

Generations

gendatedecisiontrainholdouttokensanchor LOCnote

Built 2026-09-09 22:20 UTC · therealagi.ai · a harness that writes itself, under an immutable core.