Daily digest of the Bittensor ecosystem

中文/EN

Daily edition
Nov 10, 2026

Bittensor Daily A 1T1B product
1T1B ↗
BITTENSOR DAILY

24H Market

01/10

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

THE DAILY BRIEFING

Bittensor Daily — November 10, 2026

Today's Discord traffic again centered on the stuck root-dividend deposits, with #General producing a new and more alarming variant of the complaint while #Rao stayed quiet on the issue entirely. Elsewhere, a member's own investigation into a subnet registration-fee drop surfaced a one-way pricing ratchet mechanism, and the subnet cards carried a lium.io transparency question and a SN82 governance dispute over UID control. No new protocol release shipped beyond two non-breaking PRs.

Today's highlights

  1. In #General, the root-basket deposit freeze enters a third day as a member reports their basket vanished entirely after unstaking from root.

    Same here they say is a bug but is going on for few days now sucks I then noticed that my root basket for that validator vanished I don't know what the status is of a fix. (1) This is the third consecutive day the freeze has surfaced in #General, now with a new and more alarming symptom -- a basket appearing to disappear rather than simply stall -- that the moderator could not diagnose beyond pointing to a support channel. (2) This follows Nov 7's confirmation that Root dividends were accruing but blocked from reaching stakers' baskets, and Nov 8's #Rao critique that the blocking gate does more harm than the bug it guards against. (3) Unlike Nov 8, #Rao carried none of this discussion today, leaving #General as the only channel where the issue surfaced. The moderator recommended taking no action until an announcement, with no fix timeline given.

  2. A #General member traces a subnet registration-fee drop to a specific sudo-set parameter and lays out a one-way pricing ratchet that can only decline, never recover.

    Does anyone know why the registration fee for subnets has doubled in its decline? The registration fee for sn78 is only 85.9 TAO. I got it. Sudo_set_lock_reduction_interval param was set to 28800. The price is like a slide — easy to go down, almost impossible to climb back up (1) This is a self-driven, detailed walk-through of the pricing mechanism rather than a redirect to the FAQ, with the asker explaining that once a registration slot goes unfilled the price floor drops permanently by one tier. (2) A member's pointed question Don't you get it yet? Is there anyone else who could have done something like that? about who could execute the sudo call went unanswered. (3) The floor is described as 1 TAO, with no official voice confirming the change or its rationale.

  3. On SN82 (Noah), a community member publishes on-chain evidence alleging most UIDs are controlled by multisigs tied to the owner and the Ark validator, prompting a Root emissions cut.

    (1) A second member added data describing a rapid registration of 58 of 64 slots within about two hours before emissions began, strengthening the original claim rather than contradicting it. (2) Root network emissions to the subnet have been cut, and the team says it is negotiating with moderators over what changes are required to restore them. (3) The team has since removed the NOAH deposit requirement for new miners, reimbursed initial deposits, and lowered the maximum UID registration price to 5 TAO, but emissions remain cut as of today.

  4. On SN51 (Lium), community members press the team on whether rental fees and emissions can ever net out to a value-positive case for the token, as a single UID's share of provider payouts draws scrutiny.

    (1) The discussion follows a missing Q3 report and delayed announcements flagged earlier in the thread, with a member now asking whether the large-payout UID functions as the subnet's burn address. (2) This runs alongside @taodaily_io's report that Lium renters spent $2.32M on GPU time in October, up 74% on the month, framed on X as a potential driver of value appreciation through buybacks and burns. (3) No resolution on the rental-fee-to-emissions relationship appears in the material.

  5. Other: a #Rao member is told switching validators forfeits any not-yet-credited rewards, with no guidance on timing, while Subtensor ships two non-breaking hardening fixes.

    If I switch validators now, will the rewards that haven't been credited yet still be credited later? no. You wont have any shares minted. (1) The follow-up question on when it is safe to switch went unanswered in the channel. (2) Subtensor separately merged a fix changing root dividend payout approval from an all-or-nothing batch delay to a tolerance-based check, and an SDK change erasing hashed-account signing keys from memory once unused -- both non-breaking.

Discord conversations worth following

#General

4 REPORTS

Key

The root-basket deposit freeze reported across Nov 7-8 runs into a third day, with the channel now pointing to a named technical write-up rather than offering a fix timeline

Opening ·

A member asked How does root staking even work anymore? My unclaimed rewards has actually gone down over 2 days ?, drawing an explanation that That would mean there's been a drop in the value of reward basket, doesn't mean the amount of tokens in the basket went down.

Confirmation ·

Another member said Same here they say is a bug but is going on for few days now sucks, and when a moderator asked who had said that, was pointed back to the same thread: Go up and you see me and Ivan talking about it.

Clarification ·

A separate member reported a more alarming symptom after unstaking from root -- I then noticed that my root basket for that validator vanished-- and a third asked, worried about unstaking entirely, my rewards in my basket is going down too and I'm afraid to unstake will all my rewards despair what should I do? The moderator replied With respect to the bug that Ivan mentioned; the information is here: https://s3.hippius.com/rufus/public/root-dividend-flush-price-gate.html -- I don't know what the status is of a fix., recommending take no action until there's an announcement in the <#830075335084474390> channel about a fix. I don't know when that's going to happen.

3 things to read from it

  1. This is the third consecutive day the deposit freeze has surfaced in #General, now with the moderator citing a named technical document on the 'root-dividend-flush-price-gate' rather than only the generic 'known issue' line used on Nov 7 and Nov 8.
  2. A new symptom appeared today -- a basket appearing to vanish entirely after an unstake, rather than simply not growing -- which the moderator could not diagnose beyond directing the user to another support channel.
  3. Unlike Nov 8, when #Rao carried a community critique of the gate's design trade-offs, today's #Rao traffic did not touch the freeze at all; #General is now the only channel where the issue surfaced.

In the week’s context

  • Nov 7: Subtensor confirmed Root dividends are accruing but blocked from reaching stakers' baskets, first reported the same day independently in #General and #Rao.
  • Nov 8: a #Rao member argued the deposit-blocking gate causes more harm than the minting error it was built to prevent, while #General repeated only the generic known-issue line; no fix shipped by day's end.

Where it stands

Still open; no fix timeline has been given across three days, and today's moderator reply explicitly says the status of a fix is unknown.

Notable

A member traces an apparent drop in subnet registration fees to a specific on-chain parameter change and lays out the decay mechanism in detail

Opening ·

A member asked Does anyone know why the registration fee for subnets has doubled in its decline? The registration fee for sn78 is only 85.9 TAO., and was first told Pricing for subnet registration is dynamic so its demand driven.

Investigation ·

The asker then found the specific change themselves: I got it. Sudo_set_lock_reduction_interval param was set to 28800.

Reply ·

Another member pressed on who could have made that change: Don't you get it yet? Is there anyone else who could have done something like that?

Expansion ·

The original asker then explained the pricing mechanic at length, concluding The price is like a slide — easy to go down, almost impossible to climb back up... As long as there's one gap where nobody takes over in time, the entire price level drops permanently by one step, and the floor is 1 TAO.

3 things to read from it

  1. This is the first time #General has produced a self-driven, detailed walk-through of the registration-price ratchet mechanism rather than a one-line redirect to the FAQ.
  2. The pointed question about who could set the sudo parameter was never answered, leaving open who holds the authority to change a core pricing hyperparameter and why.
  3. The exchange is a narrow mechanism explanation rather than a policy debate, but the unanswered sudo question echoes the broader governance-control questions raised by outside observers about Subtensor earlier in the week.

Where it stands

The mechanical cause was identified by the asker themselves and not disputed, but the question of who executed the sudo call and why remains unanswered in the channel.

Other

A #Rao member asks whether switching validators forfeits not-yet-credited rewards and gets no clear path forward

Opening ·

A member asked If I switch validators now, will the rewards that haven't been credited yet still be credited later? and was told no. You wont have any shares minted.

Follow-up ·

The member then asked So, what should I do now? Should I wait until all the rewards have arrived before switching validators?, which went unanswered in the material.

What to read from it

  1. Carried almost the entire 4-message #Rao channel today, with the practical follow-up question left hanging even though the initial mechanic was answered directly.

Where it stands

Unresolved; the follow-up on timing was not addressed by anyone in the channel.

Other

Other: a testnet TAO request redirected to the taoswap server, the standing scam warning repeated twice, and a Tuesday greeting filled out the rest of the day

Routine ·

A member asking how to get a testnet tao, is there still requests-for-testnet-tao channel? was redirected to the taoswap Discord, the standing scam notice ran again in full with Always ask in public channels if you have any doubt!, and one member simply posted Happy Tuesday everyone 🫡.

What to read from it

  1. Volume rose to 30 messages, the channel's busiest of the week, but the increase was driven almost entirely by the root-basket freeze thread and repeated moderator scam warnings rather than new ecosystem ground.

Where it stands

All closed through standard redirect, reminder or no reply.

#Rao

2 REPORTS

Notable

A member asks whether switching validators now forfeits rewards not yet credited, and gets a blunt no with no further guidance

Opening ·

A member asked If I switch validators now, will the rewards that haven't been credited yet still be credited later?

Reply ·

Another member answered flatly: no. You wont have any shares minted.

Follow-up ·

The asker tried to work out the practical implication: So, what should I do now? Should I wait until all the rewards have arrived before switching validators?

3 things to read from it

  1. The follow-up question went unanswered in the channel, leaving the asker without a concrete timing strategy despite getting a clear rule on the underlying mechanic.
  2. This is a plain mechanism question about validator switching and share minting, distinct from the root-basket deposit freeze dominating #General the same day, though both concern whether staking rewards actually land when expected.
  3. The answer came from a community member, not a moderator or team voice, continuing the channel's pattern of peer-to-peer mechanism explanations rather than official clarification.

Where it stands

Half-answered: the yes/no on share minting was given directly, but the practical follow-up on when it is safe to switch validators drew no reply.

Other

Other: a moderator told a member not to flood the channel with tags and stickers

Reminder ·

A moderator told the member: don't spam this channel with tags and stickers

What to read from it

  1. A routine moderation note with no further discussion.

Where it stands

Closed.

Subnets

5 REPORTS

SN58 24h +9.44%

Attune

What happened
SN58 (Attune) completed Vector Level 3, recording 98 duels and 47,040 attempts and raising top performance from 41.7% to 87.1%. The team then moved the Vector competition to Level 4, which adds cluttered-object grasping tasks, with submissions reopening at block 9,469,000. The team also announced a first Attune Community Update livestream, set for Nov 13 at 3 PM EST on Discord, covering Vector updates and the Horizon launch.
Community
Miners objected to the immediate freeze on submissions ahead of the Level 4 transition, arguing it was unfair to those mid-training, and some asked for a defined submission deadline instead. Separately, a miner questioned how a model uploaded to huggingface a week earlier was submitted within minutes of the announced Level 4 block.
Official
The team stated it would not reverse the submission freeze, citing several days of prior notice and the need for consistency, and said submissions made before the announced block would be ignored. On the pre-uploaded model, the team said the miner had simply prepared in advance.

In debate · SN58 Level 4 transition timing and fairness · 14 participants

  • A miner argued the sudden freeze on submissions penalized those who had already invested time and money in training.

    “I've invested a considerable amount of time and money in training, an”
  • The team maintained that reversing the transition after official announcement would be unfair to miners who had already adjusted their plans.

    “reversing the decision would create uncertainty and be unfair to miners who have already adjusted their plans accordingly”
  • A miner suggested the timing of a huggingface upload relative to the announced submission block implied advance knowledge, which the team attributed to preparation.

    “Block 9469000 was 10 minutes after your announcement, so someone h”

SN82 24h −5.22%

Noah

What happened
A community member published on-chain evidence alleging that most of SN82's UIDs are controlled by multisigs linked to the owner and the Ark validator, and that the Ark's holdings are overwhelmingly composed of the subnet's own token sourced largely from owner-linked miner wallets. A second community member followed with additional on-chain data describing a rapid registration of 58 of 64 slots within about two hours before emissions began. Root network emissions to SN82 have been cut, and the subnet team says it is negotiating with moderators over what changes are required to restore them. The team also announced changes to its Ark contribution model, including removal of the NOAH deposit requirement for new miners, reimbursement of initial deposits, and a lower maximum UID registration price of 5 TAO.
Community
Community members pressed the team for clarity on the update timeline and on whether the subnet's current high-incentive miners belong to the owner. A moderator described a warning referencing how another subnet was previously shut down as not a threat, prompting the team to say it still felt threatening given the moderator's role; a separate miner called the 90-day immunity ramp and new-miner emission share unfair.
Official
The team says it has made changes based on feedback, reimbursed initial NOAH deposits, prepared an update removing the Ark deposit requirement, and is seeking clarity from moderators on what is needed to restore emissions.

In debate · SN82 (Noah) self-dealing allegations and the team's response · 6 participants

  • A community member argues on-chain data shows the owner controls most UIDs and the Ark's holdings, calling the registration and emission structure risk-free self-enrichment.

    “32 sit on a multisig co-signed by 5HEZQ7Gw, which also signs the owner and Ark-validator multisigs.”
  • The subnet team maintains it has engaged with feedback, made changes, and disputes that pointing to how another subnet was shut down was not threatening given the speaker's moderator role.

    “Sharing a post about how to “kill” a subnet and applying it to ours comes across as a threat, even without explicitly saying “you’re next.””
  • A miner says the immunity ramp and emission share changes remain unfair to new miners and questions whether high-incentive miners belong to the owner.

    “If you admit the current high-incentive miners are yours or not, plz tell me here any possible way to be alive for over 45 days”

SN67 24h −0.57%

Harnyx

What happened
SN67 (Harnyx) pushed a series of repository updates: miner endpoint registration was simplified by making block_at_registration optional, validator judging for Gemma and Qwen was routed through OpenRouter, and benchmark groundwork was added to support a dual-champion mechanism alongside enforced research workspace boundaries. Validator package versions were bumped twice to reflect these changes.
Community
Several community members discussed the cost of running Terra for inference, with some arguing that API costs can exceed rewards and should be reconsidered by the project. A miner separately asked the team to raise miner emissions, citing high costs from local testing and platform validation.
Official
The team shipped repository updates covering miner endpoint registration, validator judging routes, and dual-champion benchmark groundwork, as described in the linked commits.

SN24 24h +8.98%

ms.computer

What happened
The new subnet owner, ms.computer, announced plans to launch SN24 (ms.computer) this week, saying mainnet testing had gone well and that final tweaks were being completed before onboarding more miners.
Community
Some community members questioned whether the subnet was active and asked for updates, while others expressed confidence in the new team and urged patience. A moderator reminded participants not to attempt to moderate others themselves and warned against insults in the channel.
Official
The team posted that ms.computer is launching this week, with mainnet testing complete and miner onboarding instructions to follow.

SN61 24h −0.27%

RedTeam

What happened
SN61 (RedTeam) said its CEO represented the subnet at the UK House of Lords to discuss cybersecurity threat detection technology, framing it as recognition of the subject at a government level rather than only by enterprise teams. The subnet's dashboard went down for several hours during the period and was reported multiple times by users. A miner also reported that submissions committed about 10 hours earlier had not been processed, asking whether there was a server issue.
Community
Community members discussed the House of Lords appearance, with some noting the team had gone quiet for several days beforehand after saying updates on clients and buybacks were coming. Others pointed out that many Bittensor-linked companies have been similarly quiet recently. A moderator reminded a community member to keep investor-related communication to channels outside the server.
Official
The official X account confirmed the CEO's appearance at the House of Lords, stating the subnet was introduced to members there and invited due to the relevance of its technology to recent major threats.

More subnet community coverage every day in the Telegram channel →

Ecosystem debate

12 participants

SN51 (Lium): how do rental fees, provider payouts and emissions tie into buybacks and burns?

In the SN51 (Lium) channel, community members have repeatedly asked how GPU rental fees, provider payouts and subnet emissions relate to reported buybacks and burns, after a missing Q3 report and delayed announcements were flagged earlier in the thread. The discussion intensified when a community member pointed to on-chain data showing a single UID receiving a large share of provider-pool payouts, and asked whether it functions as the subnet's burn address.

Positions

  • Some community members say the subnet should be more transparent about whether revenue is actually used for buybacks and burns, and want a way to verify it on-chain.

    “does this subnet regularly use revenue for buybacks and burns? if so, is there a dashboard where we can track them?”
  • A channel moderator's clarification distinguishes rental fees, which go to providers, from subnet emissions, which are described as a separate stream, addressing part of the confusion over the revenue-to-burn gap.

    “rental fees are 95% of what renters pay for a provider's GPUs. That 95% is paid to the provider.”
  • A community member cites on-chain data showing one UID taking a large share of provider-pool payouts and asks directly whether that address is the designated burn or treasury wallet.

    “UID 47 (5Exx2V…) receives ~72% of the provider pool, and its coldkey sends batch payouts to SN51 provider wallets daily at 17:00 UTC.”
Official response
A community manager has repeatedly acknowledged the buyback, treasury and UID-payout questions in the channel but has deferred a detailed answer to the team each time; as of the messages reviewed, no substantive response explaining the buyback/burn mechanics or the role of the high-share UID had been posted.
What to watch
Whether the team publishes a dashboard or report clarifying how rental fees, provider payouts and emissions connect to buybacks and burns, and whether it identifies the function of the UID receiving the large provider-pool share.

Voices on X

  • @Old_Samster

    Old_Samster says subnet ecosystem annualized revenue rose from roughly $20M at the end of 2025 to about $54M today, a 2.5x increase in under a year.

  • @byteleap_ai

    byteleap_ai outlines three recent protocol changes -- Null Consensus, eTAO, and Gamma -- giving subnet owners more control over scaling, incentives, and alpha burning for service credits.

  • @JesusMartinez

    JesusMartinez notes that the CEO of SN61 (RedTeam) attended an event at the House of Lords last week.

  • @Hippius_cloud

    Hippius_cloud announces its app is now available on iPhone and Android, offering automatic encrypted backup of photos and files.

  • @Old_Samster

    Old_Samster shares a chart showing 10.9% of TAO issuance currently offset by network economic activity, with a stated goal of reaching 100%.

  • @urnetwork

    urnetwork reports the winner of its SN1 (Apex) competition round, whose code cut p95 request time from 19.6 to 7.1 seconds and roughly doubled large-download speeds.

  • @taodaily_io

    taodaily_io reports SN51 (Lium) renters spent $2.32M on GPU time in October, a 74% increase from the prior month, alongside ongoing buybacks and burns.

  • @realcryptopikin

    realcryptopikin lists subnets with reported revenue, including SN64 (Chutes) at $3.69M cumulative through September 2026, SN51 (Lium) at $2.60M, SN4 (Targon) at about $1.205M, SN93 (Bitcast) at $487.5K, and SN110 (Green Compute) over $82K.

◆2026-11-10 · End of this edition

Protocol

2 PRs

#3319

fix(root): admit dividends by weighted basket price impact

What
Subtensor changed how root dividend payouts are approved, replacing a rule that delayed an entire batch if any single held subnet's price moved outside its expected range with a tolerance-based check that allows small, bounded price-impact distortions.
Why
The previous rule meant one volatile subnet price could hold up dividend settlement for an entire batch; the new approach admits payouts unless the distortion exceeds set tolerances, while subnets that do fall outside the band are rolled back individually without losing alpha rights, and non-final deferred credits with no remaining value are no longer clawed back at finalization.
Who it affects
Validators and TAO holders who receive root dividends should see more consistent and timely settlement; miners are not directly affected.

#3317

sdk: erase hashed-account Ed25519 signing keys on drop

What
The SDK was updated to erase Ed25519 signing keys for hashed accounts from memory once they are no longer in use.
Why
This closes a gap where sensitive key material could remain in memory after use, reducing exposure if that memory were later accessed.
Who it affects
This is a security hardening change relevant to developers and services building on the SDK; it does not change behavior for ordinary TAO holders, validators or miners.

Recent editions

Past editions +