Every lesson points the student toward sources, dashboards and evidence trails.
Bittensor Field College
A practical curriculum for learning Bittensor through source work, live dashboards, subnet mechanics and independent field judgment.
50 modules
50 active
5 phases
The path moves from useful work to Yuma, dTAO, emissions, slippage and thesis building.
0 modules remain in editorial review before public release.
Each module opens with the operating idea
The student sees the mechanism first, then moves into evidence, numbers and failure modes.
Every lesson ends with assessment
The quizzes are designed to test judgment, not memorized slogans or obvious answers.
Completion leads to a public learning artifact
The final path is built around progress, score, study hours and a subnet thesis style finish.
01 to 08
Protocol foundations
Useful work, miners, validators, subnets and the state surface a student must learn before reading markets.
Bittensor is a market for useful work
The first Bittensor Field College lesson: understand Bittensor as a market for measurable work before judging any subnet, chart or narrative.
Bittensor is an intelligence market
Learn why Bittensor should be read as a live market for machine intelligence, with workers, judges, incentives, capital and evidence.
Subnets are incentive machines
A subnet is closer to a living contest than a normal crypto project. Learn the parts before judging the story.
Miners are workers with an optimization problem
Miners do not mine hashes. They chase a scoring function that may or may not produce real value.
Validators are judges with capital behind them
Validators translate subnet work into weights. Their job is technical, economic and political at the same time.
Yuma turns opinions into emissions
Yuma Consensus is the on chain referee that converts validator weights into miner and validator rewards.
The weight copying problem is the hidden tax on lazy validation
Why validators copying each other weakens the measurement layer and why commit reveal exists.
Every subnet is only as good as its scoring model
A beautiful mission with a bad scoring model becomes a machine for rewarding the wrong behavior.
09 to 18
Subnet mechanics
Yuma, weights, metagraphs, hotkeys, alpha tokens and incentive design as operating machinery.
Multiple mechanisms let a subnet pay for more than one kind of work
Some subnets split emissions across different tasks. That changes how miners compete and how analysts should read performance.
The metagraph is the subnet truth table
Learn how UIDs, hotkeys, coldkeys, stake, weights and dividends appear inside a subnet state view.
Hotkeys and coldkeys explain who can act and who owns
Wallet structure matters because registration, mining, validation and ownership use different keys.
Root is the subnet for people who do not want to pick a subnet yet
Subnet zero gives subnet agnostic exposure through validators, but it has its own tradeoffs.
Staking is voting with exposure
Staking TAO into alpha pools is capital selecting which subnet deserves attention.
dTAO changed the center of gravity
Dynamic TAO moved subnet value discovery away from root judgment and toward capital flows.
Alpha tokens are claims on subnet specific belief
Each alpha token is a local market for one subnet, priced against TAO through its pool.
The subnet AMM makes every stake a price event
Bittensor subnet tokens use pools where liquidity, price and slippage shape what a position really means.
Slippage is the cost of being too large for the pool
A subnet can look cheap until your own trade moves it. Learn to think in depth, not only price.
Emissions are the payroll of the network
TAO and alpha emissions decide who gets paid, what gets reinforced and what dies quietly.
19 to 28
dTAO market structure
Pools, slippage, Taoflow, emissions, TaoSwap and the places where capital changes the field.
TaoFlow made inflow and outflow the scoreboard
The newer flow model rewards subnets that attract net TAO inflows and punishes negative flow.
Conviction is commitment with a clock attached
Conviction locks alpha to show longer term commitment and gives analysts a new ownership signal.
Subnet registration is the cost of opening a new arena
Anyone can create a subnet if they can pay and survive the rules. That is powerful and dangerous.
Deregistration is how the network makes room
Subnets and neurons can lose slots. Learn why scarcity of attention is built into the system.
Hyperparameters are the rule sheet under the scoreboard
Tempo, registration cost, UID limits and validator permits change how a subnet behaves.
A subnet GitHub can tell you what the website hides
Read repos for incentive code, update cadence, miner instructions and signs of real engineering.
A subnet without a product surface needs a harder question
Some subnets serve developers or validators only. Others should have a visible product. Learn the difference.
TAO.app is the official front door
Use TAO.app for explorer data, validator performance, tokenomics and Savant style discovery.
TaoSwap shows the market as a living table
Use TaoSwap data for market cap, flow, price evolution, holders and conviction across subnets.
TaoFlows turns staking into a tape
Use TaoFlows to watch real time liquidity movement instead of waiting for weekly summaries.
29 to 38
Research desk
TaoStats, SubnetRadar, GitHub, APIs, IntoTAO and product evidence for live subnet work.
Backprop is a screener for the dTAO market
Use Backprop for market cap, volume, top gainers, subnet categories and trading context.
SubnetRadar is for risk and anomaly hunting
Use SubnetRadar to monitor health, conviction, signals and events that deserve follow up.
AlphaGap reads what investors miss between code and market
Use AlphaGap style thinking to connect dev updates, social velocity, wallets and valuation gaps.
IntoTao is the ecosystem map for newcomers
Use IntoTao to discover resources, tools, wallets, education and project surfaces without drowning.
TaoStats is the encyclopedia with an API spine
Use TaoStats docs, explorer, staking views and API references to cross check almost everything.
APIs turn Bittensor from story into evidence
Learn which APIs expose subnets, metagraphs, holders, transactions, slippage and social data.
Subnet categories are useful until they become lazy
LLM, compute, data, agents, media and finance subnets need different evaluation criteria.
LLM subnets compete on routing, quality and cost
Understand what to inspect in inference, model development and language related subnets.
Compute subnets sell capacity and reliability
Read GPU, serverless and infrastructure subnets through uptime, demand, utilization and pricing.
Data subnets live or die by provenance and demand
Scraping, storage and data generation subnets need buyers, quality controls and audit trails.
39 to 50
Field judgment
Selection, red flags, yield traps, conviction, wallet analysis and the final subnet thesis.
Agent subnets should be judged by tasks completed
Agents are narrative magnets. Learn to ask what the agent actually does, for whom and how often.
Media subnets turn attention into a measurable commodity
Bitcast, video, content and verification subnets show how Bittensor can reward media production.
Financial subnets need risk language, not hype language
Lending, trading, liquidity and prediction subnets must be read through counterparties and failure modes.
The red flags that make a subnet hard to trust
No repo, thin docs, unclear scoring, fake volume, missing product and bad incentives deserve attention.
A fragile subnet can still be honest
Learn the difference between malicious behavior, weak design, early experimentation and honest failure.
Validator selection is delegated judgment
Choosing a validator means choosing how your stake is represented across work markets.
Yield without context is the easiest trap in TAO
APY, emissions, alpha price and slippage can point in different directions.
Wallet flows reveal behavior before explanations do
Track accumulation, rotation, concentration and exits across subnet markets.
GitHub research is a habit more than a metric
Commits, releases, issues and docs should be read as evidence of shipping quality.
How to write a subnet thesis without fooling yourself
Turn product, incentives, code, flows and risk into a thesis you can update.
Build your Bittensor research terminal
Assemble the daily desk: docs, GitHub, TAO.app, TaoFlows, TaoSwap, Backprop, SubnetRadar, AlphaGap and your notes.
The subnet thesis exam
The final module gives a repeatable exam for judging any subnet from zero.
The field method
This is not passive onboarding. The student learns to read claims, test them against source data, calculate the market cost of being wrong and build a subnet thesis that can survive pressure.
- ReadUnderstand the mechanism before judging the narrative.
- VerifyOpen the source, dashboard, repo or API behind the claim.
- CalculateTranslate price, APY, liquidity, flow and slippage into actual risk.
- DecideWrite the thesis, name the invalidation and keep the evidence trail.