Bittensor Toolkit →Find the right tool for your question.

BITTENSOR · THE NEWS ARCHIVE

Find the story.

Search 86 dated editions without scrolling through them one by one.

Reset filters

Dates and verification labels belong to the original edition. A past verified announcement can now be outdated. Historical project names are retained; an SN filter searches the slot across registrations. All keywords must match the story text or subnet label. Ratings describe editorial priority, not investment quality.

2,335 matching stories

2026-10-27 edition →SN4 · Targonverified4 / 5 stars

Targon reports revenue-funded buybacks and burns; a large alpha burn is finalized

SN4 Targon’s October 27 announcement reports repurchasing and burning 1,772 TAO-equivalent of alpha using organic Targon.com revenue, plus another 6,678 TAO-equivalent of alpha described as revenue-linked. Its graphic combines those as 8,450 TAO; these are valuations of alpha, not 8,450 native TAO destroyed. Independently, our finalized scan confirms 152,314.359019724 SN4 alpha burned at block 9,376,324, extrinsic 7, at 16:44:36 UTC. That verifies token destruction, not the claimed revenue provenance, repurchase cost or full TAO-equivalent breakdown. The distinction matters when assessing whether customer revenue is supporting subnet token demand.

A consequential revenue-linked capital-allocation announcement supported by a separately verified large alpha burn; revenue attribution and TAO valuations remain explicitly team-reported.

2026-10-27 edition →SN8 · Vantaverified3 / 5 stars

Vanta records a qualifying 103.438971428-alpha transaction burn

A successful SN8 AlphaBurned event at block 9,373,978 destroyed 103.438971428 alpha on October 27 at 08:55:24 UTC. The event is retained in the public burn archive with its account, transaction and historical registration label. This is a subnet burn by the recorded wallet; it is not automatically an owner-wallet burn. No additional qualifying burn was found in the subsequent fully scanned increment through the daily cutoff.

A verified transaction burn above the individual 100-alpha threshold; no emissions withholding or pooled cross-subnet totals substituted.

2026-10-27 edition →SN17 · 404—GENdeveloping2 / 5 stars

404—GEN delays its competition wrap-up while API work continues

The October 27 brief says the final Competition 3 update is delayed while API work remains unfinished, with a further update expected rather than completed. This is a schedule/operations item, not a product launch or independently validated model result.

A routine competition delay, kept below the major-news threshold.

2026-10-27 edition →SN22 · Desearchverified3 / 5 stars

Desearch’s Reddit demo returns public discussion threads

Desearch announced its Reddit Search demo on October 27. Our Bittensor topic test returned ten public thread links with subreddit names and titles. The published example uses one web search restricted to Reddit; it does not read complete threads or generate an analysis, and dates may be absent. Alongside its October 26 RAG extraction example, it shows developers a small, reproducible discovery workflow rather than a claim of fresh or authoritative Reddit coverage.

A working, inspected developer tool with limited scope; no unsupported adoption claim.

2026-10-27 edition →SN33 · ReadyAIverified3 / 5 stars

ReadyAI shifts its allocation toward validators

The finalized October 27 SubnetEmissionWeightsSet event at block 9,376,469, extrinsic 23, changes SN33 ReadyAI’s displayed owner/validator/miner split from approximately 18%/41%/41% to 18%/49%/33%. Exact five-slot vectors and registration identity are preserved in the evidence; the raw owner weight changes slightly despite the rounded 18% label. A larger validator allocation is not a guaranteed individual validator or staker APY, and historical yield measurements can lag the change.

Verified executed allocation change affecting reward distribution, without projecting returns.

2026-10-27 edition →SN34 · Yugosverified3 / 5 stars

SN34’s finalized identity is now Yugos

At finalized block 9,375,531, SN34’s owner-submitted identity reads Yugos and points to the BitMind-AI/yugos repository, describing decision models for inference. This is a verified identity observation, not proof of an ownership sale or a new registration. Older BitMind stories keep their historical label. The current brief and code also discuss evaluation and sandbox restoration; these should be assessed as current Yugos work.

A verified identity clarification that prevents readers carrying an older product label into current coverage.

2026-10-27 edition →SN36 · Bittensorverified2 / 5 stars

SN36 changes its owner-submitted display name

The finalized identity at block 9,375,531 names SN36 “Bittensor,” replacing the Epago label seen in earlier coverage. Its accompanying description is owner-submitted commentary, not an assessment endorsed by this publication. A name edit alone does not establish a sale, new registration, partnership or change to the Bittensor network itself.

Identity housekeeping, with the potentially confusing display name explicitly separated from the network.

2026-10-27 edition →SN44 · Scoredeveloping3 / 5 stars

Score removes problematic evaluation cases and reports an outage

Score’s October 27 brief describes blacklist/ground-truth corrections in its petrol-PPE evaluation and a Beverage service interruption. A published blacklist commit supports ongoing maintenance, but this review did not reproduce the incident or audit scores. Miners should use the current operator guidance; removed cases do not establish a general model-performance gain.

Operational relevance for participants; competition and incident details stay attributed.

2026-10-27 edition →SN49 · Nepher Roboticsverified3 / 5 stars

Nepher reports Phase 1 navigation results in simulation

Nepher’s October 27 update reports 76 submissions from 27 participants in its Unitree Go2 LiDAR challenge and a best reported score of 0.9272. The evaluation concerns unseen mazes in Isaac Lab. These are competition results in simulation, not a demonstrated production deployment on physical robots; the practical question is how later stages generalize beyond that test.

Routine competition results capped at three stars, with simulation and physical deployment kept distinct.

2026-10-27 edition →SN50 · Synthdeveloping3 / 5 stars

Synth adds Bybit perpetual trade data to its Ultra feed

The October 27 operator brief says Synth Ultra adds a 60-second Bybit USDT perpetual trade stream under other_venues.bybit_futures.trades, with schema version 6. Consumers need to check schema handling rather than assume the existing parser covers the new venue. The brief says miner burn settings are unchanged; that is not an on-chain transaction-burn announcement.

An actionable data/schema update, attributed to the operator brief.

2026-10-27 edition →SN54 · Yanezverified3 / 5 stars

Today’s Shizzy interview: Yanez and action-bound identity

The Yanez interview with Jose Caldera is scheduled for October 27 at 12:30 PM ET / 9:30 AM PT. Yanez’s October 26 explanation says Pulse approval binds a specific action to the owner’s live-face verification and that moment. The stream is upcoming at this report’s cutoff, not a replay or independent security audit.

Shizzy video/interview coverage is three stars under the owner’s current rubric.

2026-10-27 edition →SN54 · Yanezverified3 / 5 stars

Yanez executes its new split and explains the end of miner withholding

The finalized October 27 event at block 9,376,080, extrinsic 12, changes SN54 Yanez’s displayed owner/validator/miner split from approximately 18%/41%/41% to 18%/49%/33%; exact raw vectors are archived. The team says its prior 30% miner burn ends and describes higher liquidity-pool TAO inflow and alpha-staker yield at then-current stake. Those economic estimates are team calculations, not fixed returns or an independently measured outcome. Miner withholding is separate from an explicit transaction burn and is not added to our Burns archive. Root’s baseline/cap and historical yield lag still apply.

Executed allocation change with clearly separated team economic estimates and transaction-burn accounting.

2026-10-27 edition →SN55 · NIOMEdeveloping3 / 5 stars

NIOME schedules the first stage of its GENOME migration

NIOME’s October 26 post puts the first 5% stage of its GENOME-on-Base to locked SN55 NIOME migration two days later, October 28. Treat this as a scheduled migration, not completed token delivery. Separately, the October 27 brief reports task-generator recovery and testnet work. Holders should distinguish the announced schedule from an executed, independently checked migration.

A dated holder-facing plan and operational update; execution remains unconfirmed.

2026-10-27 edition →SN64 · Chutesdeveloping4 / 5 stars

Chutes outlines off-subnet inference and a broader compute roadmap

Jon Durbin’s October 26 roadmap separates some additional inference options, which he says are not funded by subnet emissions, from the subnet’s incentive work. Proposed next steps include private TEE jobs, hourly GPU/CPU compute and training or serving on participants’ hardware. He explicitly says secure TEE inference and GPUs continue. These are plans and strategic direction, not evidence that all listed products have launched or become profitable.

A consequential operating and incentive strategy from the project lead; clearly labelled as a roadmap.

2026-10-27 edition →SN64 · Chutesverified3 / 5 stars

Chutes expands the non-TEE model catalog

SN64 Chutes’ October 27 follow-up announces seven additional non-TEE model choices under its existing API key and billing, alongside its TEE offerings. This is an incremental availability update to the previously covered beta: users gain more model choices, but non-TEE availability should not be read as carrying the same confidential-compute properties as TEE inference. We reviewed the team post and model-family artwork; no paid inference benchmark or independent availability test was performed.

A useful catalog expansion rather than a newly verified adoption or performance breakthrough.

2026-10-27 edition →SN67 · Harnyxverified4 / 5 stars

Harnyx connects SN67 research miners to REAX decision models

Harnyx announced on October 27 that its research miners can now call decision models from REAX, SN92. Its published integration adds Fast and Pro providers for the smaller routing and stopping decisions inside a research run. This is a concrete connection between two subnet products; it does not establish the number of paying users, actual cost savings or independently measured research quality.

A documented cross-subnet product integration with an immediately usable role, rather than a competition result.

2026-10-27 edition →SN71 · Leadpoetverified3 / 5 stars

Leadpoet interview remains scheduled for October 28

Leadpoet’s interview is still scheduled for October 28 at 11 AM ET / 8 AM PT. Its watch page is an upcoming broadcast, so it stays out of completed-video coverage. Current repository changes tighten arena readiness/release checks; those commits alone do not establish a new product launch.

A verified upcoming Shizzy stream; no premature replay or launch claim.

2026-10-27 edition →SN73 · Elenostverified3 / 5 stars

SN73 changes from Parked to Elenost and adopts a 33% / 6% / 61% split

SN73’s owner-submitted identity changed from Parked to Elenost. A separate successful allocation event at block 9,372,736 changes owner/validator/miner proportions from 18%/41%/41% to 33%/6%/61%. The archive preserves the exact five-slot raw vectors, block hash and extrinsic. These are allocations, not a promise of a particular validator APY. Root baseline/cap treatment and observed yield can differ or lag; the split is not an alpha burn.

A verified economic allocation change; historical labels and exact chain evidence preserved.

2026-10-27 edition →SN78 · Umideveloping3 / 5 stars

Umi describes a C6 rerun with a larger hidden evaluation set

Umi’s October 27 brief describes rerunning C6 against a larger hidden set after concerns about evaluation coverage and memorization. Published code also bounds retained reward coverage at successor activation. The reported future emission increase is a plan, not verified new yield. No individual miner is accused here, and evaluation concerns are not treated as proof of misconduct.

A material evaluation/process correction, with allegations and planned economics carefully qualified.

2026-10-27 edition →SN90 · KubeTEEverified3 / 5 stars

KubeTEE publishes its attestation specification and records a model withdrawal

KubeTEE’s published repository changes add an attestation specification/schema and record withdrawal of GLM-5.3-Flash. Related documentation clarifies publication status and key-distribution assumptions. This is evidence of published engineering material, not a finding that all east–west security goals are deployed or independently audited. Operators should check the actual supported model list before relying on an older availability claim.

A substantive compatibility and security-documentation update, without turning a specification into a completed deployment.

Searches published headlines and summaries, not private bookmarks, unpublished leads or source-page contents. Duplicate headlines within an edition appear once; separate editions remain separate so follow-ups are preserved. Share the page URL to retain your filters.