Skip to content

docs: ban non-causal TokenMix label-leak architectures #7

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
echobt wants to merge 1 commit into from
Closed

docs: ban non-causal TokenMix label-leak architectures #7

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 11 additions & 0 deletions docs/scoring.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,17 @@ submission is coherent and not cheating; their quality notes never move the scor
- Probe the cheap gate first with `POST /v1/submissions/precheck` (3/coldkey/UTC
day) — see [Submit](submit.md#precheck-similarity-before-you-submit).
Comment on lines +20 to +21

@coderabbitai coderabbitai Bot Aug 8, 2026

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🗄️ Data Integrity & Integration | 🟡 Minor | ⚡ Quick win

State the quota as three requests per coldkey per UTC day.

3/coldkey/UTC day does not identify the counted unit clearly. Use the full phrase from the PR contract so clients know that the limit applies to precheck requests.

🤖 Prompt for AI Agents 
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@docs/scoring.md` around lines 20 - 21, Update the quota wording in the
precheck guidance near the Submit reference to explicitly state that the limit
is three precheck requests per coldkey per UTC day, replacing the ambiguous
shorthand while preserving the endpoint and link.



## Causal LM contract (banned: non-causal label leak)

Prism scores **next-token** cross-entropy → BPB. Architectures must not let
position `t` read tokens `t+1…` (including the label). Dense sequence mixers —
MLP-Mixer-style `TokenMix` / `t_mix` / `nn.Linear` over the full time axis after
`transpose(1, 2)` — **without** a causal mask (`triu` / `tril` / `is_causal` /
attention mask) are a hard ban (`non_causal_label_leak`, `Score(0)`, terminal,
often caught **before** GPU rent). Channel mixing and causal attention / causal
conv are fine; bidirectional full-sequence mixes used as a next-token LM are not.
Comment thread
coderabbitai[bot] marked this conversation as resolved.

## Architecture competition (emission math)

Per epoch, your emission is the **max** of:
Expand Down
Loading