{ "netuid": 40, "slug": "chunking", "name": "Chunking", "symbol": "\u05df", "description": "Chunking Bittensor subnet", "priceTao": 0.003463626, "priceUsd": 0.8764743890887332, "athUsd": 0.8764743890887332, "change24h": 0.21410636920155104, "change7d": 0.984295013274571, "change30d": 2.1937441469766092, "change90d": 0.0, "marketCapUsd": 3508547.9439842673, "volume24hUsd": 12486.08925000082, "rootProp": 0.15726468846287983, "liquidityUsd": 3600973.49263551, "alphaStaked": 1948535.380366861, "activeMiners": 1, "activeValidators": 13, "emissionPct": 0.00721241692451348, "emissionPerDayTao": 1.0, "registrationCost": 0.0, "pruningRank": 80, "immunityRemaining": 14265, "sentiment": "Neutral", "githubRepo": "https://github.com/VectorChat/chunking_subnet", "subnetUrl": "", "scorecard": { "adoption": { "label": "mixed", "score": 49, "reasons": [ { "tone": "neutral", "label": "Stars", "value": "0", "weight": null }, { "tone": "neutral", "label": "Contributors", "value": "0", "weight": null }, { "tone": "positive", "label": "24h volume", "value": "$798,336,535", "weight": null }, { "tone": "neutral", "label": "Validators", "value": "11", "weight": null } ] }, "shipSpeed": { "label": "weak", "score": 0, "reasons": [ { "tone": "neutral", "label": "Commits 30d", "value": "0", "weight": null }, { "tone": "neutral", "label": "Merged PRs 30d", "value": "0", "weight": null }, { "tone": "neutral", "label": "Releases 12m", "value": "0", "weight": null }, { "tone": "negative", "label": "Latest push", "value": "Unknown", "weight": null } ] }, "updatedAt": "2026-03-09T20:06:23.088076+00:00", "confidence": 0.45, "usefulness": { "label": "watch", "score": 50, "reasons": [ { "tone": "neutral", "label": "Durable keywords", "value": "0 matched", "weight": null }, { "tone": "positive", "label": "Commodity risk", "value": "0 matched", "weight": null }, { "tone": "neutral", "label": "Freshness", "value": "Unknown", "weight": null }, { "tone": "positive", "label": "Subnet traction", "value": "20/20", "weight": null } ] }, "codeQuality": { "label": "weak", "score": 0, "reasons": [ { "tone": "negative", "label": "CI", "value": "Missing", "weight": null }, { "tone": "negative", "label": "Tests", "value": "Missing", "weight": null }, { "tone": "neutral", "label": "Security hygiene", "value": "No SECURITY.md", "weight": null }, { "tone": "neutral", "label": "Issue closure", "value": "0 issues / 30d", "weight": null } ] } }, "marketScore": 82.0, "githubScore": 0.0, "aiScore": 50.0, "masterScore": 41.0, "masterRating": "avoid", "lastUpdatedAt": "2026-04-29T20:21:48Z", "isStale": false, "purpose": { "title": "Chunking Bittensor subnet", "source": "fallback", "confidence": 0.42, "shortSummary": "Chunking Bittensor subnet" }, "repoCoverageStatus": "disabled", "repos": [], "repoSnapshots": [], "analysis": null, "externalLinks": [ { "label": "GitHub", "href": "https://github.com/VectorChat/chunking_subnet" } ], "priceHistory": [ { "timestamp": "2025-12-26T20:59:48-03:00", "value": 0.9666168 }, { "timestamp": "2025-12-27T20:59:48-03:00", "value": 0.97356727 }, { "timestamp": "2025-12-28T20:59:48-03:00", "value": 0.97337343 }, { "timestamp": "2025-12-29T20:59:48-03:00", "value": 0.99546537 }, { "timestamp": "2025-12-30T20:59:48.001000-03:00", "value": 0.99128745 }, { "timestamp": "2025-12-31T20:59:48.001000-03:00", "value": 0.99243786 }, { "timestamp": "2026-01-01T20:59:48.001000-03:00", "value": 0.99008663 }, { "timestamp": "2026-01-02T20:59:48.001000-03:00", "value": 0.99417364 }, { "timestamp": "2026-01-03T20:59:48-03:00", "value": 1.03128938 }, { "timestamp": "2026-01-04T20:59:48-03:00", "value": 1.0418968 }, { "timestamp": "2026-01-05T20:59:48.001000-03:00", "value": 1.07343443 }, { "timestamp": "2026-01-06T20:59:48-03:00", "value": 1.07941717 }, { "timestamp": "2026-01-07T20:59:48-03:00", "value": 1.07929745 }, { "timestamp": "2026-01-08T20:59:48-03:00", "value": 1.09589208 }, { "timestamp": "2026-01-09T20:59:48-03:00", "value": 1.11324473 }, { "timestamp": "2026-01-10T20:59:48-03:00", "value": 1.11467789 }, { "timestamp": "2026-01-11T20:59:48-03:00", "value": 1.08307142 }, { "timestamp": "2026-01-12T20:59:48.001000-03:00", "value": 1.0962613 }, { "timestamp": "2026-01-13T20:59:48-03:00", "value": 1.17505177 }, { "timestamp": "2026-01-14T20:59:48-03:00", "value": 1.16505228 }, { "timestamp": "2026-01-15T20:59:48.001000-03:00", "value": 1.17825983 }, { "timestamp": "2026-01-16T20:59:48.001000-03:00", "value": 1.18525863 }, { "timestamp": "2026-01-17T20:59:48-03:00", "value": 1.20723449 }, { "timestamp": "2026-01-18T20:59:48.001000-03:00", "value": 1.24542418 }, { "timestamp": "2026-01-19T20:59:48-03:00", "value": 1.26892166 }, { "timestamp": "2026-01-20T20:59:48.001000-03:00", "value": 1.26583159 }, { "timestamp": "2026-01-21T20:59:48-03:00", "value": 1.33130684 }, { "timestamp": "2026-01-22T20:59:48-03:00", "value": 1.29728028 }, { "timestamp": "2026-01-23T20:59:48-03:00", "value": 1.3069685 }, { "timestamp": "2026-01-24T20:59:48-03:00", "value": 1.26171425 }, { "timestamp": "2026-01-25T20:59:48-03:00", "value": 1.24341245 }, { "timestamp": "2026-01-26T20:59:48-03:00", "value": 1.31300763 }, { "timestamp": "2026-01-27T20:59:48-03:00", "value": 1.30080268 }, { "timestamp": "2026-01-28T20:59:48-03:00", "value": 1.24483645 }, { "timestamp": "2026-01-29T20:59:48-03:00", "value": 1.26707955 }, { "timestamp": "2026-01-30T20:59:48-03:00", "value": 1.2332279 }, { "timestamp": "2026-01-31T20:59:48.001000-03:00", "value": 1.26273115 }, { "timestamp": "2026-02-01T20:59:48.001000-03:00", "value": 1.26336463 }, { "timestamp": "2026-02-02T20:59:48-03:00", "value": 1.25322757 }, { "timestamp": "2026-02-03T20:59:48-03:00", "value": 1.24911128 }, { "timestamp": "2026-02-04T20:59:48.001000-03:00", "value": 1.2588868 }, { "timestamp": "2026-02-05T20:59:48.001000-03:00", "value": 1.29968335 }, { "timestamp": "2026-02-06T20:59:48.001000-03:00", "value": 1.28633216 }, { "timestamp": "2026-02-07T20:59:48-03:00", "value": 1.26201906 }, { "timestamp": "2026-02-08T20:59:48.001000-03:00", "value": 1.26986275 }, { "timestamp": "2026-02-09T20:59:48-03:00", "value": 1.28413365 }, { "timestamp": "2026-02-10T20:59:48.001000-03:00", "value": 1.29696713 }, { "timestamp": "2026-02-11T20:59:48-03:00", "value": 1.31435527 }, { "timestamp": "2026-02-12T20:59:48-03:00", "value": 1.29849599 }, { "timestamp": "2026-02-13T20:59:48-03:00", "value": 1.29879657 }, { "timestamp": "2026-02-14T20:59:48.001000-03:00", "value": 1.30746404 }, { "timestamp": "2026-02-15T20:59:48-03:00", "value": 1.29049065 }, { "timestamp": "2026-02-16T20:59:48-03:00", "value": 1.29405909 }, { "timestamp": "2026-02-17T20:59:48-03:00", "value": 1.30217602 }, { "timestamp": "2026-02-18T20:59:48.001000-03:00", "value": 1.32221281 }, { "timestamp": "2026-02-19T20:59:48-03:00", "value": 1.32502407 }, { "timestamp": "2026-02-20T20:59:48.001000-03:00", "value": 1.32770586 }, { "timestamp": "2026-02-21T20:59:48-03:00", "value": 1.33528469 }, { "timestamp": "2026-02-22T20:59:48-03:00", "value": 1.33774993 }, { "timestamp": "2026-02-23T20:59:48-03:00", "value": 1.31794137 }, { "timestamp": "2026-02-24T20:59:48-03:00", "value": 1.31781992 }, { "timestamp": "2026-02-25T20:59:48-03:00", "value": 1.30078678 }, { "timestamp": "2026-02-26T20:59:48-03:00", "value": 1.29521004 }, { "timestamp": "2026-02-27T20:59:48-03:00", "value": 1.31647499 }, { "timestamp": "2026-02-28T20:59:48-03:00", "value": 1.30872732 }, { "timestamp": "2026-03-01T20:59:48-03:00", "value": 1.33338495 }, { "timestamp": "2026-03-02T20:59:48.001000-03:00", "value": 1.34442525 }, { "timestamp": "2026-03-03T20:59:48-03:00", "value": 1.36386187 }, { "timestamp": "2026-03-04T20:59:48-03:00", "value": 1.36822099 }, { "timestamp": "2026-03-05T20:59:48-03:00", "value": 1.36907701 }, { "timestamp": "2026-03-06T20:59:48.001000-03:00", "value": 1.36629296 }, { "timestamp": "2026-03-07T20:59:48-03:00", "value": 1.37435909 }, { "timestamp": "2026-03-08T20:59:48.001000-03:00", "value": 1.36608525 }, { "timestamp": "2026-03-09T13:50:00-03:00", "value": 0.76960625 }, { "timestamp": "2026-03-09T19:24:36-03:00", "value": 0.76883597 }, { "timestamp": "2026-03-09T19:25:48-03:00", "value": 0.76883636 }, { "timestamp": "2026-03-09T19:42:12-03:00", "value": 0.76884063 }, { "timestamp": "2026-03-09T19:54:48-03:00", "value": 0.76884238 }, { "timestamp": "2026-03-09T20:59:48-03:00", "value": 1.36464406 }, { "timestamp": "2026-03-10T20:59:48-03:00", "value": 1.35604037 }, { "timestamp": "2026-03-10T22:04:12-03:00", "value": 0.78983065 }, { "timestamp": "2026-03-11T04:57:36.001000-03:00", "value": 0.75432986 }, { "timestamp": "2026-03-11T20:59:48-03:00", "value": 1.34423997 }, { "timestamp": "2026-03-12T20:59:48.001000-03:00", "value": 1.330948 }, { "timestamp": "2026-03-13T20:59:48-03:00", "value": 1.29780088 }, { "timestamp": "2026-03-14T20:59:48.001000-03:00", "value": 1.30399936 }, { "timestamp": "2026-03-15T09:16:36-03:00", "value": 1.00572754 }, { "timestamp": "2026-03-15T20:59:48-03:00", "value": 1.29955157 }, { "timestamp": "2026-03-16T15:03:36-03:00", "value": 1.08017551 }, { "timestamp": "2026-03-16T20:59:48-03:00", "value": 1.27649039 } ], "liquidityHistory": [ { "timestamp": "2025-12-26T20:59:48-03:00", "value": 3776700.04451808 }, { "timestamp": "2025-12-27T20:59:48-03:00", "value": 3797249.51763372 }, { "timestamp": "2025-12-28T20:59:48-03:00", "value": 3803861.53914454 }, { "timestamp": "2025-12-29T20:59:48-03:00", "value": 3853892.49737021 }, { "timestamp": "2025-12-30T20:59:48.001000-03:00", "value": 3852917.02463337 }, { "timestamp": "2025-12-31T20:59:48.001000-03:00", "value": 3862298.09251608 }, { "timestamp": "2026-01-01T20:59:48.001000-03:00", "value": 3864848.1231304 }, { "timestamp": "2026-01-02T20:59:48.001000-03:00", "value": 3879971.82108949 }, { "timestamp": "2026-01-03T20:59:48-03:00", "value": 3959122.19130249 }, { "timestamp": "2026-01-04T20:59:48-03:00", "value": 3986922.8634002 }, { "timestamp": "2026-01-05T20:59:48.001000-03:00", "value": 4054490.85931359 }, { "timestamp": "2026-01-06T20:59:48-03:00", "value": 4073549.13026512 }, { "timestamp": "2026-01-07T20:59:48-03:00", "value": 4081042.16917461 }, { "timestamp": "2026-01-08T20:59:48-03:00", "value": 4120143.31466756 }, { "timestamp": "2026-01-09T20:59:48-03:00", "value": 4160599.47565291 }, { "timestamp": "2026-01-10T20:59:48-03:00", "value": 4171309.45661003 }, { "timestamp": "2026-01-11T20:59:48-03:00", "value": 4119590.77407263 }, { "timestamp": "2026-01-12T20:59:48.001000-03:00", "value": 4152480.56088923 }, { "timestamp": "2026-01-13T20:59:48-03:00", "value": 4307402.56328292 }, { "timestamp": "2026-01-14T20:59:48-03:00", "value": 4297467.41710616 }, { "timestamp": "2026-01-15T20:59:48.001000-03:00", "value": 4330225.4633107 }, { "timestamp": "2026-01-16T20:59:48.001000-03:00", "value": 4351594.27589295 }, { "timestamp": "2026-01-17T20:59:48-03:00", "value": 4400405.78743007 }, { "timestamp": "2026-01-18T20:59:48.001000-03:00", "value": 4478371.42738416 }, { "timestamp": "2026-01-19T20:59:48-03:00", "value": 4529502.89762819 }, { "timestamp": "2026-01-20T20:59:48.001000-03:00", "value": 4533082.64349905 }, { "timestamp": "2026-01-21T20:59:48-03:00", "value": 4658308.38144197 }, { "timestamp": "2026-01-22T20:59:48-03:00", "value": 4607748.26048852 }, { "timestamp": "2026-01-23T20:59:48-03:00", "value": 4634322.53533376 }, { "timestamp": "2026-01-24T20:59:48-03:00", "value": 4562495.30905291 }, { "timestamp": "2026-01-25T20:59:48-03:00", "value": 4538247.50090223 }, { "timestamp": "2026-01-26T20:59:48-03:00", "value": 4672915.90008077 }, { "timestamp": "2026-01-27T20:59:48-03:00", "value": 4660523.08058927 }, { "timestamp": "2026-01-28T20:59:48-03:00", "value": 4568243.95702163 }, { "timestamp": "2026-01-29T20:59:48-03:00", "value": 4617958.66039895 }, { "timestamp": "2026-01-30T20:59:48-03:00", "value": 4564762.62346973 }, { "timestamp": "2026-01-31T20:59:48.001000-03:00", "value": 4628048.79475712 }, { "timestamp": "2026-02-01T20:59:48.001000-03:00", "value": 4638290.98475744 }, { "timestamp": "2026-02-02T20:59:48-03:00", "value": 4628155.17315974 }, { "timestamp": "2026-02-03T20:59:48-03:00", "value": 4629516.10545561 }, { "timestamp": "2026-02-04T20:59:48.001000-03:00", "value": 4656664.20022541 }, { "timestamp": "2026-02-05T20:59:48.001000-03:00", "value": 4740851.09278058 }, { "timestamp": "2026-02-06T20:59:48.001000-03:00", "value": 4725735.03143689 }, { "timestamp": "2026-02-07T20:59:48-03:00", "value": 4689998.74801872 }, { "timestamp": "2026-02-08T20:59:48.001000-03:00", "value": 4713694.66346966 }, { "timestamp": "2026-02-09T20:59:48-03:00", "value": 4749335.92327483 }, { "timestamp": "2026-02-10T20:59:48.001000-03:00", "value": 4782311.60267507 }, { "timestamp": "2026-02-11T20:59:48-03:00", "value": 4823707.81230102 }, { "timestamp": "2026-02-12T20:59:48-03:00", "value": 4803893.23885468 }, { "timestamp": "2026-02-13T20:59:48-03:00", "value": 4813758.14842344 }, { "timestamp": "2026-02-14T20:59:48.001000-03:00", "value": 4839211.42451698 }, { "timestamp": "2026-02-15T20:59:48-03:00", "value": 4817017.87658748 }, { "timestamp": "2026-02-16T20:59:48-03:00", "value": 4832996.2174459 }, { "timestamp": "2026-02-17T20:59:48-03:00", "value": 4857504.24886869 }, { "timestamp": "2026-02-18T20:59:48.001000-03:00", "value": 4904233.73004023 }, { "timestamp": "2026-02-19T20:59:48-03:00", "value": 4918826.25543388 }, { "timestamp": "2026-02-20T20:59:48.001000-03:00", "value": 4933228.42110154 }, { "timestamp": "2026-02-21T20:59:48-03:00", "value": 4956903.4510749 }, { "timestamp": "2026-02-22T20:59:48-03:00", "value": 4971109.89057911 }, { "timestamp": "2026-02-23T20:59:48-03:00", "value": 4943583.51429307 }, { "timestamp": "2026-02-24T20:59:48-03:00", "value": 4952784.06759302 }, { "timestamp": "2026-02-25T20:59:48-03:00", "value": 4929494.17824663 }, { "timestamp": "2026-02-26T20:59:48-03:00", "value": 4927996.27954411 }, { "timestamp": "2026-02-27T20:59:48-03:00", "value": 4977612.64508153 }, { "timestamp": "2026-02-28T20:59:48-03:00", "value": 4972115.72407642 }, { "timestamp": "2026-03-01T20:59:48-03:00", "value": 5028068.83590385 }, { "timestamp": "2026-03-02T20:59:48.001000-03:00", "value": 5058504.20550009 }, { "timestamp": "2026-03-03T20:59:48-03:00", "value": 5104733.06259141 }, { "timestamp": "2026-03-04T20:59:48-03:00", "value": 5122738.76745773 }, { "timestamp": "2026-03-05T20:59:48-03:00", "value": 5134199.8094034 }, { "timestamp": "2026-03-06T20:59:48.001000-03:00", "value": 5138786.7015944 }, { "timestamp": "2026-03-07T20:59:48-03:00", "value": 5163774.02680602 }, { "timestamp": "2026-03-08T20:59:48.001000-03:00", "value": 5158056.91413016 }, { "timestamp": "2026-03-09T13:50:00-03:00", "value": 2909679.85201786 }, { "timestamp": "2026-03-09T19:24:36-03:00", "value": 2909509.99869999 }, { "timestamp": "2026-03-09T19:25:48-03:00", "value": 2909515.34737585 }, { "timestamp": "2026-03-09T19:42:12-03:00", "value": 2909586.48445462 }, { "timestamp": "2026-03-09T19:54:48-03:00", "value": 2909638.23203111 }, { "timestamp": "2026-03-09T20:59:48-03:00", "value": 5165159.97535438 }, { "timestamp": "2026-03-10T20:59:48-03:00", "value": 5158625.05564518 }, { "timestamp": "2026-03-10T22:04:12-03:00", "value": 3004852.3768457 }, { "timestamp": "2026-03-11T04:57:36.001000-03:00", "value": 2889279.8383385 }, { "timestamp": "2026-03-11T20:59:48-03:00", "value": 5145756.1421495 }, { "timestamp": "2026-03-12T20:59:48.001000-03:00", "value": 5129487.3282984 }, { "timestamp": "2026-03-13T20:59:48-03:00", "value": 5072668.67710033 }, { "timestamp": "2026-03-14T20:59:48.001000-03:00", "value": 5090934.60038383 }, { "timestamp": "2026-03-15T09:16:36-03:00", "value": 3911457.57672538 }, { "timestamp": "2026-03-15T20:59:48-03:00", "value": 5087755.65772051 }, { "timestamp": "2026-03-16T15:03:36-03:00", "value": 4206366.32148514 }, { "timestamp": "2026-03-16T20:59:48-03:00", "value": 5047169.57188915 } ], "scoreHistory": [ { "date": "2026-03-09", "masterScore": 41.0, "marketScore": 82.0, "githubScore": 0.0, "aiScore": 50.0, "socialScore": 0.0, "rating": "avoid", "usefulness": 50, "codeQuality": 0, "adoption": 49, "shipSpeed": 0 } ], "githubHistory": [], "socialAccounts": [ { "handle": "chunking_subnet", "role": "project", "label": "Official", "source": "manual", "confidence": 1.0, "profile_image_url": "https://pbs.twimg.com/profile_images/1820841888775036928/wp0MBcO-_normal.jpg" } ], "recentTweets": [ { "tweet_id": "1872071536980717998", "handle": "chunking_subnet", "text": "Patch 2.4.2 - Commit Reveal\nPR: https://t.co/DfM6rogYHe\n\nKey Changes:\n- Enable CR3\n- Chunks now properly end on sentence boundary, with NLTK tokenizer as ground truth\n- When miners tie in a round, alpha now decreases linearly instead of exponentially", "created_at": "2024-12-25T21:06:50-03:00", "likes": 13, "retweets": 3, "replies": 3, "views": 1611, "is_retweet": false, "is_reply": false, "media_type": null }, { "tweet_id": "1867265400901161451", "handle": "chunking_subnet", "text": "Release 2.4.0 - Enhanced Competition & Variety\nPR: https://t.co/lF1NDWxb4f\n\nKey Changes:\n- New 'Loss Alpha' system accelerates rank advancement for consistently better miners\n- Expanded chunk size variety: 2000-4000 chars with varying probabilities\n- Mixed prompt types: 80% complex interwoven, 20% clear separation\n- Added rank protection for winning miners\n\nLoss Alpha Multipliers:\nGroup 0: 2x\nGroup 1: 1.25x\nGroup 2: 1.0625x\nGroup 3: 1.015625x\n\nAs we surpass industry standards, this system ensures better miners can advance faster despite increasingly gradual improvements.", "created_at": "2024-12-12T14:48:58-03:00", "likes": 6, "retweets": 0, "replies": 1, "views": 1380, "is_retweet": false, "is_reply": false, "media_type": null }, { "tweet_id": "1866548921247420827", "handle": "chunking_subnet", "text": "Latest Chunking Subnet Benchmarks\nThe Power of Open, Incentivized Contest\nCompleted Dec 9, 2024\n\nWe have now surpassed the industry leaders @AI21Labs and @UnstructuredIO in intermediate contexts, and have widened our lead in low context!\n\n\ud83d\udd17 Explore the interactive benchmark here: https://t.co/BHgaZLxf3G\n\nKey findings:\n\ud83d\udca0 The subnet outperformed industry leaders at low (50 tokens) and intermediate (150) context\n\ud83d\udca0 The subnet widened its lead in low context to +25.99% on recall and +3.92% on precision\n\ud83d\udca0 The subnet is now winning on recall +8.93% and precision +7.66%\n\ud83d\udca0 The top miners did significantly better than the top miner of the last benchmark!\n\nConsiderations\n\ud83d\udd38 This benchmark follows the same methodology as the previous benchmark, apart from the following:\n\ud83d\udd38 New questions were generated using the process outlined by @trychroma here: https://t.co/kJfape62J7\n\ud83d\udd38 The dataset now includes documents from openwebtext and wikitext\n\ud83d\udd38 In the time since the last benchmark, AI21 released a new chunker", "created_at": "2024-12-10T15:21:56-03:00", "likes": 40, "retweets": 12, "replies": 3, "views": 6973, "is_retweet": false, "is_reply": false, "media_type": null }, { "tweet_id": "1859338905427853371", "handle": "chunking_subnet", "text": "Release 2.2.0 - Enhanced Query Generation\n\nPR: https://t.co/M8Ye6GQFMt\n\nKey Changes:\n- Improved system, initial generation, and continuation prompts\n- Synthetic queries now weave together topics from each of three given articles much more, often even within the same sentence\n\nExplore examples between the old and new generation: https://t.co/Tqk9VY82Dc", "created_at": "2024-11-20T17:51:54-03:00", "likes": 23, "retweets": 4, "replies": 0, "views": 2004, "is_retweet": false, "is_reply": false, "media_type": null }, { "tweet_id": "1851935703334010915", "handle": "chunking_subnet", "text": "Patch 2.1.2\n\nPR: https://t.co/3kMBbjEmMf\n\nKey Changes:\n- When creating a unique hash for all chunks returned by a miner, whitespace is now ignored.\n- Extra checks to make sure nltk tokenizers are downloaded", "created_at": "2024-10-31T07:34:13-03:00", "likes": 8, "retweets": 2, "replies": 0, "views": 789, "is_retweet": false, "is_reply": false, "media_type": null }, { "tweet_id": "1850987492822171911", "handle": "chunking_subnet", "text": "Hotfix 2.1.1\n\nPR: https://t.co/nN8Ts88KX3\n\nKey Changes\n- Improved Validator Task API concurrency.\n- Improved W&B tracking to address the copy-source feature on https://t.co/7Y3kyDnla9, with additional classifications for rounds.\n\nValidators\n- Make sure to restrict access to the Task API to only allow in trusted traffic\n- We recommend a cron restart of 1 hour\n\nMiners\n- Ensure your miner can handle concurrent requests, as we'll likely be sending in a significant amount of requests as part of the commercialization benchmark", "created_at": "2024-10-28T16:46:22-03:00", "likes": 9, "retweets": 1, "replies": 0, "views": 562, "is_retweet": false, "is_reply": false, "media_type": null }, { "tweet_id": "1849482527829795322", "handle": "chunking_subnet", "text": "Release 2.1.0 - Task API & https://t.co/xk22bV1GMy Network\n\nPR: https://t.co/mVHcTq2I42\n\nKey Changes:\n- New integrated task API enabling concurrent organic & synthetic queries\n- Sidecar API server for validators to handle client requests\n- Removed sleep time between synthetic tournament rounds\n- Optimized validator performance with async functions\n\nFor Commercialization Benchmark:\n- Enable with --enable_task_api\n- Whitelist https://t.co/xk22bV1GMy (IP in Discord)\n\nRemember, the commercialization benchmark starts tomorrow and runs until November 1st.\n\n1 TAO base reward, 50 TAO reward pool based on % queries answered.\n\nSee full implementation details and setup guide: https://t.co/wMX9fOc4co", "created_at": "2024-10-24T13:06:11-03:00", "likes": 25, "retweets": 6, "replies": 3, "views": 7568, "is_retweet": false, "is_reply": false, "media_type": null }, { "tweet_id": "1845945307634585676", "handle": "chunking_subnet", "text": "Patch 1.4.1 - Ranking System Improvements\n\nSee PR: https://t.co/d7fR9zW7Yv\n\nThis patch addresses two key issues in the current ranking system that made it challenging for newer miners to advance.\n\nIssue 1: Identical Chunks, Inconsistent Rewards\n- OpenAI\u2019s Embedding models are non-deterministic, meaning the same text can generate slightly different embeddings with each submission.\n- As a result, miners submitting identical chunks could receive slightly different scores, leading to unequal rewards for the same work.\n\nResolution:\n- Now, if multiple miners submit the same chunk, validators will evaluate the chunk only once, ensuring consistent rewards for all miners submitting identical work.\n\nIssue 2: Two Groups, One Ranking.\n- Many miners are assigned to two groups: one where they can improve their rank and another that could drag their rank down based on the performance of other miners.\n- Because miner ranks are calculated using a weighted moving average, some miners may experience rank drops despite performing well in their lower group.\n\nExample: If a miner ranks 2.7, they could be in both Group 1 (miners ranked 1-4) and Group 2 (ranked 3-8). Even if they consistently rank 1st in Group 2, their overall rank could still fall because 3 is higher than their initial rank of 2.7.\n\nResolution:\nThe starting rank value of each group will now be based on the average of two values:\n- The rank at the overlapping position in the adjacent higher group.\n- The last rank of the second adjacent higher group.\n\nFrom this derived starting point, ranks will increment by 1 for each subsequent position, creating a more balanced ranking system.\n\nFor a more detailed explanation of the issues, solutions, and the logic behind these changes, please refer to our documentation here: https://t.co/2ZPsUXvZy0", "created_at": "2024-10-14T18:50:32-03:00", "likes": 18, "retweets": 3, "replies": 1, "views": 1793, "is_retweet": false, "is_reply": false, "media_type": null }, { "tweet_id": "1843653095504351728", "handle": "chunking_subnet", "text": "Testnet Release 1.4.0\n\nPR: https://t.co/9rYoUh00hq\n\nChangelog (diff to mainnet):\n1. Synthetic query upgrade: gpt-4o-mini generates unique 15k-60k char articles from Wikipedia\n2. Weight hanging fix: Only waits for inclusion, not finalization\n3. Reward function fix: Addresses tokenizer punctuation edge-cases\n\nUpcoming:\n- Lucid anti-relay system on testnet post-merge\n- Mainnet merge: Wed 16:00 UTC\n- Lucid mainnet: Oct 16th\n\nNew: You can now copy source text from dashboard rounds at https://t.co/3xFhFI09CT\n\nLucid docs: https://t.co/YxAd5J2qcZ", "created_at": "2024-10-08T11:02:06-03:00", "likes": 26, "retweets": 6, "replies": 2, "views": 2439, "is_retweet": false, "is_reply": false, "media_type": null }, { "tweet_id": "1831410522711101840", "handle": "chunking_subnet", "text": "Patch 1.3.1\n\nView the PR here: https://t.co/uIDTMSDd7h \n\nChangelog:\n\n1. Incentives are no longer cutoff after the first 5 ranks.\n- The first seven ranks will continue to follow the function (1/2)^N\n- The remaining incentive (1/128) is divided linearly (with a nearly zero, but non-zero slope) according to rank for ranks 8-200\n\n2. The min burn parameter will rise to 100,000,000 or 0.1 TAO.\n\n3. API framework is merged into main.\n\n4. Adds custom verify function to the Default Miner, to handle \"Nonce too old\" error causing miners to refuse some validator queries.\n\nThis adds a new option \n--neuron.synapse_verify_allowed_delta\nso miners can choose when to deem validator synapses as stale. The default is now 10 seconds.\n\nOf course, verification can always be disabled completely using\n--neuron.disable_verification,\nthough this is not recommended.", "created_at": "2024-09-04T15:14:29-04:00", "likes": 21, "retweets": 3, "replies": 2, "views": 2225, "is_retweet": false, "is_reply": false, "media_type": null }, { "tweet_id": "1829570505956901260", "handle": "chunking_subnet", "text": "Hotfix 1.2.3\n\nThis hotfix address an exploit which allowed miners to bypass restrictions on data integrity. The PR can be found here: https://t.co/gdaWxjsITq \n\nNote that the dashboard will lag for a few hours as the wandb API catches up. If you need to see the latest info immediately, please use our wandb page: https://t.co/zJfBDdtpow \n\nValidators not using auto-update, please update as soon as possible with\n\ngit checkout main\ngit pull\npip3 install -e .\nbash https://t.co/vOeBzKPwiA", "created_at": "2024-08-30T13:22:55-04:00", "likes": 5, "retweets": 1, "replies": 0, "views": 362, "is_retweet": false, "is_reply": false, "media_type": null }, { "tweet_id": "1826337340253753580", "handle": "chunking_subnet", "text": "Patch 1.2.1\n\nCorrects group size pattern from 4, 6, 8, 10 to 2, 4, 6, 8, 10\n\nView PR here: https://t.co/m06mynbEeR\n\nNew Groups by Rank:\n[[0, 1], [1, 2, 3, 4], [3, 4, 5, 6, 7, 8], [6, 7, 8, 9, 10, 11, 12, 13], [10, 11, 12, 13, 14, 15, 16, 17, 18, 19], [15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26], [21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34], [28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43], [36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53], [45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64], [55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76], [66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89], [78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103], [91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118], [105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134], [120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151], [136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169], [153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199]]", "created_at": "2024-08-21T15:15:28-04:00", "likes": 1, "retweets": 0, "replies": 2, "views": 284, "is_retweet": false, "is_reply": false, "media_type": null }, { "tweet_id": "1826331413378941015", "handle": "chunking_subnet", "text": "Release 1.2.0 Live on Mainnet\n\nThis is a breaking change for Validators with auto-update without a WandB key.\n\nSee the full details on our PR: https://t.co/8EOZ9gIyAA\n\nChangelog:\n1. Groups now having varying alphas, with lower ranked groups more sensitive to change and the higher ranked groups less sensitive to change.\n\n2. Group sizes are no longer fixed, decreasing in size as ranks rise. Group sizes follow the pattern 4, 6, 8, 10 with overlap between each group increasing by 1 each time.\n\n3. Fixed \"all rankings\" in WandB logs\n\n4. Increased default num_embeddings to 150\n\n5. WandB API key is now required in the start script\n\nFor Validators, please update at your earliest convenience.\n\nNote that a Weights & Biases API Key is now required for the validator start script. Please set this variable in your .env.", "created_at": "2024-08-21T14:51:55-04:00", "likes": 5, "retweets": 1, "replies": 0, "views": 1003, "is_retweet": false, "is_reply": false, "media_type": null }, { "tweet_id": "1822794007702900788", "handle": "chunking_subnet", "text": "\ud83d\ude80 Patch 1.1.0 on Testnet!\n\nKey updates:\n\n\ud83e\udde9 Evaluation changes (see quoted tweet for details)\n\n\ud83d\udcb0 Incentive mechanism adjustments\n\n- Temporarily limiting rewards to top 5 miners to achieve consensus sooner \n- Will revert to top 17 once sufficient differentiable miners outperform the default\n\n\ud83d\udcca Enhanced Wandb logging\n\nChunking Dashboard v2 launches soon after the patch! New logging features to support the upcoming dashboard include:\n\n- Global benchmark every few hours via https://t.co/xk22bV1GMy Task API, showing how each miner handled the same document\n- For each validator, all synthetic queries made\n- For each miner, all synthetic queries answered\n- For each synthetic query, the miner group formed, the document provided, how each miner chunked the response, similarity score given, and relative ranking within the group.\n\nGoes live on finney Aug 14, 16:00 UTC.", "created_at": "2024-08-11T20:35:31-04:00", "likes": 27, "retweets": 4, "replies": 6, "views": 2275, "is_retweet": false, "is_reply": false, "media_type": null }, { "tweet_id": "1822788823304020091", "handle": "chunking_subnet", "text": "Testnet Patch 1.1.0: Evaluation Update\n\nWe're excited to announce important changes to our evaluation mechanism for the chunking subnet. This article explains the rationale behind these changes, how they're implemented, and what they mean for our subnet.\n\nCurrent System (v1.0.5)\nThe current system works as follows:\n\n1. Forms \"small chunks\" of 3 sentences from each miner chunk\n2. Randomly selects a sample (default 50) to embed\n3. Calculates dot product of every possible pair of embeddings\n4. Adds pairs from same chunk to score (rewarding intrachunk similarity)\n5. Subtracts pairs from different chunks (penalizing interchunk similarity)\n\nThe Issue\nA flaw of this design is that it encouraged creating less chunks. With fewer chunks, there's less likelihood of the random sample finding \"small chunks\" from different chunks. Since the dot product of pairs between chunks was always subtracted, having less of these pairs increased your score.\n\nThis was originally mitigated by the `chunk_size` attribute, which exponentially penalized exceeding the specified chunk size.\n\nHowever, this shifted the focus from optimizing for the most semantically meaningful chunks to optimizing how much to exceed `chunk_size` to create fewer chunks.\n\nOur Solution (v1.1.0)\nInstead of considering each pair individually, we now use averages:\n\nIntrachunk similarity = \u03a3(same chunk pairs) / n(same chunk pairs) \n\nInterchunk similarity = \u03a3(different chunk pairs) / n(different chunk pairs)\n\nReward = Intrachunk similarity - Interchunk similarity - Penalties\n\nThis new system weighs intrachunk and interchunk similarity equally, regardless of the number of chunks created.\n\nTo control for the size and number of chunks, validators have two new parameters: `min_chunk_size` and `max_chunk_size,` both of which will still have exponential penalties.\n\nExpected Outcomes\nOur local testing shows this system more accurately rewards semantically meaningful responses. We're excited to see how it performs in the wider subnet environment!\n\nDiagram of the Evaluation Process, as of v1.1.0:", "created_at": "2024-08-11T20:14:55-04:00", "likes": 18, "retweets": 3, "replies": 2, "views": 2803, "is_retweet": false, "is_reply": false, "media_type": null } ] }