Daily digest of the Bittensor ecosystem

中文/EN

Daily edition
Nov 8, 2026

Bittensor Daily A 1T1B product
1T1B ↗
BITTENSOR DAILY

24H Market

01/10

Gainers
--Loading…--
Losers
--Loading…--

THE DAILY BRIEFING

Bittensor Daily — November 8, 2026

Both ecosystem channels stayed light, but the day's main thread was a protocol-level friction point surfacing twice: a blocked deposit path for Root earnings in #Rao and an unclaimed-basket display problem in #General, both attributed to a recent upgrade without a confirmed fix date. Subnet news carried more weight than usual, from REAX's claimed world-leading decision model to a transparency complaint on RedTeam's alpha buybacks and a code-exposure fix under consideration on Ditto. No tokenomics or governance debate of the kind seen earlier in the week resurfaced in either channel.

Today's highlights

  1. A #Rao member asks why Root earnings aren't reflected in their basket, and the team confirms a deposit block that a separate #General thread independently describes the same day.

    Root generated earnings today, so why hasn't the amount in the basket changed? The team replied There was a change that currently blocks the deposits (fix being worked on), the dividends are still accruing and pending though. The same day in #General, a member reported Since yesterday my "unclaimed basket" remains 0 on https://taostats.io/pro/stake/claim (also on Talisman i don't get any unclaimed balance anymore). Is this a known issue?, and another member explained Your basket is growing but you do not see it - effect from most recent upgrade. (1) Two channels describe what looks like the same underlying problem without cross-referencing each other, suggesting the issue is wider than either thread alone shows. (2) This follows the week's run of breaking staking and basket changes -- Release 487's bundle and Release 488's reversal on Nov 5 -- though nothing in either thread ties the block explicitly to those releases. (3) Neither thread got a timeline for a fix or confirmation of how the backlog will be credited once resolved.

  2. A #Rao member presses on when third-party wallets like Polkadot.js will support quantum-resistant keys, and the team can only point to its own extension in progress.

    Is polkadot js (or any other than btcli) suppprting quantum resistant wallets anytime soon? drew the reply Not sure how that would work/look like PolkadotJS extension or other wallets etc. all I could say is that we are already working on the support for them in the TAOApp Extension 🫡. (1) This narrows the third-party-wallet gap flagged on Nov 6, when a member detailed that hardware wallets and root-level activation lagged the native rollout, to one specific provider. (2) The team gave no timeline for Polkadot.js support itself, offering only its own in-house extension as a concrete answer. (3) The asker closed the exchange asking to be notified when it ships, with no further detail provided.

  3. SN118 (Ditto)'s team proposes moving evaluations off validators after acknowledging that validators running dittobench currently receive the full agent docker image, exposing Python agent code.

    (1) The team flagged that because Python is interpreted, unlike Rust, validators holding the full docker image can see submitted agents' code -- a disclosed design weakness rather than an alleged exploit. (2) Its proposed fix pairs this with a plan to drop automated screening in favor of harder benchmark difficulty and manual daily review, continuing the subnet's pattern of near-weekly emission and scoring changes seen since Nov 1's tokenomics rework. (3) The team is considering, but has not committed to, moving evaluations to private infrastructure with replayable receipts.

  4. SN92 (REAX) says it has trained the world's best decision model and announces a new target it calls building the world's most intelligent model.

    (1) The team confirmed the milestone in its own channel after an external post praised the result, framing it as reached in under two weeks. (2) This follows REAX 1.3's Nov 6 leaderboard result, where both Opentensor and an independent account credited the subnet on the same day -- today's announcement extends that attention into a stated next goal rather than a new independent benchmark. (3) No methodology or independent verification of either claim appears in the material.

  5. A SN61 (RedTeam) miner publicly questions InnerWorks' alpha buyback pace, noting only a small amount has been repurchased in over three months despite a revenue-share pledge.

    (1) The miner called for transparent, consistent buybacks reflecting the pledged revenue share and suggested the allocation could eventually rise. (2) The comparison to another subnet team's higher revenue commitment frames this as a relative transparency gap rather than a specific accusation. (3) The team had not responded in the material by day's end.

  6. Other: #General's day stayed routine, with FAQ redirects, a retired self-assignable role, and a REAX staking question filling most of the traffic.

    A member asking the channel to check an unfamiliar subnet query was told We don't get involved in subnet queries we don't know, while another asking about staking was pointed to a channel and replied Yes, indeed, thank you 🙂. (1) Volume held close to Nov 6's level, continuing a run of support-dominated days. (2) None of the remaining traffic touched tokenomics, governance or protocol direction.

Discord conversations worth following

#General

2 REPORTS

Notable

A member reports their unclaimed stake basket stuck at zero since yesterday, and two different answers surface in the channel about whether it's a known issue

Opening ·

A member reported Since yesterday my "unclaimed basket" remains 0 on https://taostats.io/pro/stake/claim (also on Talisman i don't get any unclaimed balance anymore). Is this a known issue?

Reply ·

One moderator said they had no information on it: I'm not aware of any issue. Maybe ask in the Taostats Discord.

Clarification ·

A second reply contradicted that, calling it Yes known issue and explaining the mechanism: Your basket is growing but you do not see it - effect from most recent upgrade.

3 things to read from it

  1. Two moderator-style replies to the same question diverge -- one denies any known issue, the other confirms it and attributes it to a recent upgrade -- leaving the asker with conflicting information rather than a single authoritative answer.
  2. The same friction surfaced the same day in #Rao, where a member asked why root earnings weren't reflected in their basket and was told deposits are currently blocked by a change the team is fixing, though dividends keep accruing -- the two channels describe what looks like the same underlying problem without cross-referencing each other.
  3. This follows a week of breaking protocol changes to staking and basket mechanics: Release 487's bundle and Release 488's reversal on Nov 5, and Gamma being pushed to chain but left switched off on Nov 6 -- any of which could plausibly explain a basket-display anomaly, though nothing in the channel ties the fix explicitly to those releases.

In the week’s context

  • On Nov 5, Subtensor shipped Release 488, reverting the staking, basket, proxy and EVM fee and gas-pricing changes bundled into Release 487 just days earlier.
  • On Nov 6, Gamma was pushed to the Bittensor chain but remained switched off pending root-wallet activation, per a Taostats notice.

Where it stands

Unresolved in the channel: the second reply offered an explanation (growing but not displayed) but no timeline for a fix, and the first reply's denial of any known issue was never withdrawn or reconciled.

Other

Other: routine support filled most of the day, including a REAX staking question, a retired self-assignable role, a TAO-acquisition guide request, and a subnet-query redirect

Support ·

A member asked Where may I get infos about staking REAX ? Thx you ! and, after being pointed to a channel, said May I ask you for some help about REAX subnet, since there is no specific forum about it ?; after the link was reposted, they thanked the moderator with Yes, indeed, thank you 🙂

Support ·

A newcomer asked hi [mention] how can i get TAO i'm seeking for guide and was told to Read that section of the FAQ

Support ·

A member asking about a retired role, how can i get gang sign? 🙂, was told Hi there, "Gang Signs" is no longer available. and pointed to You can find all self-assignable roles in <#1179129432410173541>

Routing ·

A member asking the channel to check an unfamiliar subnet query, gm, can u check this msg - https://discord.com/channels/799672011265015819/1233428828479819876/1568526936588689540 Maybe u have some info, was told We don't get involved in subnet queries we don't know

What to read from it

  1. Volume held at 21 messages, close to Nov 6's 22, continuing a run of routine-support-dominated days; none of today's remaining traffic touched tokenomics, governance or protocol direction.

Where it stands

All items closed through standard redirect or FAQ pointer; nothing substantive remains open.

#Rao

2 REPORTS

Notable

A member asks why Root dividends aren't showing in their basket, and the team confirms a block on deposits that is independently reported the same day in #General

Opening ·

A member asked Root generated earnings today, so why hasn't the amount in the basket changed?

Clarification ·

The team replied There was a change that currently blocks the deposits (fix being worked on), the dividends are still accruing and pending though, prompting the follow-up When will the fix be completed, and once it is, will the missing amount be credited to my basket balance in a single lump sum?, which went unanswered.

Reaction ·

A regular in the channel endorsed the approach, writing imo letting dividends keep accruing while deposits are blocked is the right call and freezing the whole path would be worse, then argued for how the backlog should be released: when the fix lands id rather see the pending fill in chunks than one giant overnight pop because wallet jumps like that always look scarier than they are.

3 things to read from it

  1. The same issue surfaced independently in #General the same day, where a member reported their unclaimed basket stuck at zero and another member confirmed it as a known effect of the most recent upgrade -- corroborating #Rao's account of a deposit block without any coordination between the two threads.
  2. The question of when missing dividends get credited, and whether in one lump sum or gradually, went unanswered by the team; the only view offered on the mechanics came from a community member rather than an official source.
  3. This follows the week's run of breaking protocol changes to staking and basket functionality -- Release 487's changes and Release 488's partial reversal -- without the material confirming which specific change is responsible for the current block.

In the week’s context

  • On Nov 5, Release 488 reverted staking, basket, proxy and EVM fee and gas-pricing changes that Release 487 had introduced only days earlier, with no community reaction logged in the material at the time.

Where it stands

Open; the team confirmed dividends are accruing but deposits are blocked pending a fix, with no timeline or detail on how the backlog will be credited once it lands.

Other

A member asks whether third-party wallets like Polkadot.js will support quantum-resistant keys, and the team points only to its own extension

Question ·

A member asked Is polkadot js (or any other than btcli) suppprting quantum resistant wallets anytime soon?

Reply ·

The team said Not sure how that would work/look like PolkadotJS extension or other wallets etc. all I could say is that we are already working on the support for them in the TAOApp Extension 🫡, drawing the request Let's us know when it goes live.

What to read from it

  1. This continues the gap Nov 6's #Rao thread flagged -- that hardware and third-party wallet support lags the native rollout -- with the team today unable to say when or whether Polkadot.js itself will support the new keys, only that its own TAOApp Extension is in progress.

In the week’s context

  • On Nov 6, a #Rao member explained in detail that hardware wallets were the real bottleneck for quantum-resistant signing and that root had not yet enabled hashed accounts for native transactions; today's exchange narrows that same gap to one specific third-party wallet without resolving it.

Where it stands

Open; the team confirmed work on its own extension but gave no answer on Polkadot.js or a timeline for either.

Subnets

5 REPORTS

SN120 24h +0.29%

Affine

What happened
SN120 (Affine) worked through a multi-hour stuck evaluation pipeline on turn E824 caused by a GPU memory failure during a switch to 64K training, resolved by reverting training to 32K while keeping the 64K promotion gate; turn E834 then produced the first GREEN head since E824. A subsequent turn, E836, incorrectly promoted a new head via a timeout rather than a true GREEN score, requiring a rollback to the E834 head and a gate patch so lifecycle timeouts now fail closed. The team then introduced parallel evaluation starting at turn e843, letting the trainer move to the next turn while candidates are judged on separate eval nodes instead of a single serial gate, and later moved evaluation onto rented B300 GPUs with a warm node pool to speed up decisions, producing new GREEN heads at e843 and e861.
Community
Community members raised concerns about evaluation speed and clarity, including a question about why one submission's evaluation reportedly took up to three hours, and asked how the parallel-evaluation queue and head-replacement mechanics would affect turns already in progress.
Official
The team posted step-by-step updates on each incident, raised the probe deadline from 40 to 80 minutes, rolled back the E836 head, patched the gate to fail closed on lifecycle timeouts, and adjusted registration burn parameters to match the new parallel-evaluation throughput.

SN70 24h +35.02%

Fugal

What happened
In the SN70 (Fugal) channel, several miners questioned why their emissions had not increased and asked whether rewards were being directed outside the normal miner pool. The team stated that no such practice occurs without it being announced, and said its current priority is the development of a v2 release. No release date for v2 was given beyond assurances that it would be announced when ready.
Community
Community members were divided: some pointed to flat emissions as a concern, while others defended the team and asked for evidence before drawing conclusions. Several participants urged patience, noting the team's focus on v2, and some miners said they plan to continue on v1 in the meantime.
Official
The team denied any undisclosed allocation of rewards and said v2 has priority over all other work, with no fixed release date provided.

SN118 24h −4.21%

Ditto

What happened
SN118 (Ditto)'s team proposed removing the automated screening step for submitted agents, replacing it with harder benchmark difficulty and a manual daily review process. The team also raised a separate issue: validators running dittobench receive the full agent docker image, which can expose Python agent code since Python is interpreted, unlike Rust. To address this, the team is considering moving evaluations off validators onto private infrastructure with replayable receipts.
Community
Several miners reported submissions stuck in "Under review" status for extended periods, in some cases over a day, and asked the team to check specific agent IDs. One community member engaged with the screening-removal proposal, raising questions about how benchmark data would stay unseen before races and how cheating would be detected without automated screening.
Official
The team outlined a plan to manually review top-performing submissions daily and scale benchmark difficulty in response, and proposed moving evaluations to private infrastructure with replayable receipts to prevent Python code exposure.

In debate · SN118 (Ditto): removing automated screening in favor of manual review · 4 participants

  • The team argued removing screening cuts costs, eliminates wait times, and lets effort go toward scaling benchmark difficulty instead.

    “Miners, how do we feel about removing the screening step? A couple of advantages: cost savings, no wait time for screening”
  • A community member questioned how difficult, never-before-seen benchmark data could be produced and protected without automated screening.

    “The data should be never released before and during the race, how will you detect if someone hacked or got a dataset in special way?”
  • The team proposed moving evaluations to private infrastructure with replayable receipts to address Python code leakage via validator-run docker images.

    “I suggest we move evals off the validator and onto my private infra and have receipts so that runs can be reproduced after a delay”

SN92 24h +9.78%

REAX

What happened
An external X post praised the SN92 (REAX) team for reaching what it called world state-of-the-art on a decision model in under two weeks. The team confirmed the milestone in the subnet channel and stated plans to move on to building what it described as the world's most intelligent model, referred to internally as "GPT 10.0 Bomboclaat - Super intelligence v1.0."
Community
A community member asked how to select a validator for staking REAX, noting that Taomarketcap only supports token swaps. Another community member explained that validator selection is available on taostats.io and pointed to the subnet's own validator (UID 3); the original asker later confirmed the staking was completed and thanked the subnet team.
Official
The team stated it had trained the world's best decision model and announced its next goal of building the world's most intelligent model, under the working name "GPT 10.0 Bomboclaat - Super intelligence v1.0."

SN61 24h +0.17%

RedTeam

What happened
A miner in the SN61 (RedTeam) channel raised concerns about InnerWorks' alpha buyback practices, noting that only 115 TAO had been bought back over more than three months despite the team's pledge to allocate 10% of SaaS revenue to buybacks. The miner called for transparent, consistent, and preferably monthly buybacks that reflect that commitment, and suggested the allocation could eventually be increased beyond 10%. The same miner compared the pledge to another subnet team's commitment of 20% of revenue.
Community
One community member responded approvingly, calling the post 'well said.' Separately, other members in the channel expressed frustration over delayed team announcements and unanswered questions, describing the team's silence on related topics as giving off 'weak team feels.'

In debate · SN61 (RedTeam) alpha buyback commitment dispute · 5 participants

  • A miner argued InnerWorks' buybacks are too small relative to its pledged revenue share and should support subnet value.

    “A portion of revenue needs to be used to purchase alpha and subsequently burn or lock it.”
  • The same miner called for more transparent, monthly buybacks that reflect the company's actual business development.

    “I'd love to see transparent, consistent, and preferably monthly buybacks that actually reflect InnerWorks' business development”
  • Another community member criticized the team's silence on related questions as a sign of weak communication.

    “it's the silence/dodging of all related questions. Not a good look ; gives out weak team feels”

More subnet community coverage every day in the Telegram channel →

Voices on X

  • @opentensor

    SN64 (Chutes) partnered with Tencent Cloud and is a launch partner for Tencent Hunyuan's new Hy4 model, now available through its API.

  • @JesusMartinez

    @lium_io reported October figures of $2.32M spent on GPU time, up 74% month-over-month, with about 5% captured as revenue, $116,000 before costs and promotional credits.

  • @BitcoinNews

    A Bittensor co-founder told @ExploitSummit, in an interview shared by @BitcoinNews, why open-source AI could challenge the largest AI labs' control over compute.

  • @BitcoinNews

    The same co-founder warned, per @BitcoinNews, that compute, capital and power are concentrating in a few hands, describing this concentration as a "cabal."

  • @theminos_ai

    The SN107 (Minos) app entered public beta and is now downloadable from its website after incorporating feedback from early testers.

  • @ShizzyUnchained

    SN15 (ORO)'s new SMB Bench found eight of nine tested AI models finished below a simple fixed buying rule over a 90-day simulated produce-buying test.

  • @opentensor

    The latest Novelty Search episode highlighted advances from Macrocosmos, SN23 (Trishool), Reliquary and ChronoLLM, with teams from ChronoLLM, REAX and Reliquary joining to discuss their research.

  • @desearch_ai

    Desearch MCP added OAuth support for its Meta Muse connector, letting Muse register dynamically while tool calls still require user authorization, without a custom auth bridge.

◆2026-11-08 · End of this edition

Recent editions

Past editions +