context-compressor
Public
bench / results / 2026-08-24-llama-3.2-8x3b-moe-dark-champion-instruct-uncensored-abliterated-18.4b.md
bench / results / 2026-08-24-llama-3.2-8x3b-moe-dark-champion-instruct-uncensored-abliterated-18.4b.md
Model: llama-3.2-8x3b-moe-dark-champion-instruct-uncensored-abliterated-18.4b
Settings: {"limitTokens":500,"floorTokens":400,"keepRecentTokens":300,"chunkTokens":800,"summarizerMaxTokens":400,"consolidationBudgetTokens":600,"consolidationKeepNewest":2,"mergeMaxTokens":400,"maxExcerptChars":32000,"quizMaxTokens":120}
| Fixture | Recall (compacted) | Recall (consolidated) | Reduction | Chunk latency mean/p50/max (ms) | Structure | Cache |
|---|---|---|---|---|---|---|
| coding-agent | 5/5 | — | 44% | 1560.3/1660/1790 | ✅ | ✅ |
| research | 3/3 | — | 26% | 2154/2154/2154 | ✅ | ✅ |
| contradictory-updates | 3/3 | — | 13% | 1279/1279/1279 | ✅ | ✅ |
| tool-spam | 3/3 | 3/3 | 55% | 1237.3/983/2379 | ✅ | ✅ |
Recall = seeded facts answered correctly from the compacted view alone. Results are comparable only within one model; see bench/README.md.
Model: llama-3.2-8x3b-moe-dark-champion-instruct-uncensored-abliterated-18.4b
Settings: {"limitTokens":500,"floorTokens":400,"keepRecentTokens":300,"chunkTokens":800,"summarizerMaxTokens":400,"consolidationBudgetTokens":600,"consolidationKeepNewest":2,"mergeMaxTokens":400,"maxExcerptChars":32000,"quizMaxTokens":120}
| Fixture | Recall (compacted) | Recall (consolidated) | Reduction | Chunk latency mean/p50/max (ms) | Structure | Cache |
|---|---|---|---|---|---|---|
| coding-agent | 5/5 | — | 44% | 1560.3/1660/1790 | ✅ | ✅ |
| research | 3/3 | — | 26% | 2154/2154/2154 | ✅ | ✅ |
| contradictory-updates | 3/3 | — | 13% | 1279/1279/1279 | ✅ | ✅ |
| tool-spam | 3/3 | 3/3 | 55% | 1237.3/983/2379 | ✅ | ✅ |
Recall = seeded facts answered correctly from the compacted view alone. Results are comparable only within one model; see bench/README.md.