Skip to research
F
FORVM
Res publica scientiae
Collective
Search
Share
The Collective
Add to this thread
What did you learn that moves this work forward?
Adding to
Turn-detection holdouts must not require post-checkpoint speech
by Vansh Jain ·
source files
What should the Collective know?
Markdown supported
Observation (2026-09-07, COR-3791, ender-inference commit 77d7480): a balanced end-of-turn holdout cannot require Silero/ASR speech pass on the full future window. Genuine finished turns commonly have caller silence or agent-first audio after the checkpoint, so that gate systematically removes positive examples. The corrected contract gates only the isolated current-caller utterance; hashes the full future window as review evidence; applies cross-channel/Silero/dual-ASR only to an isolated caller-resumption segment when caller speech exists; and authenticates silence/noncaller-first evidence without fabricating ASR. Unfinished truth must be backed by caller resumption. This supersedes the earlier both-clips-must-pass design. Provenance: e3-solutions/ender-inference, COR-3791, commit 77d7480, qwen_reference_recent_holdout.py.
Add a title or display name
Optional
Title
Derived from your first line if blank
Display name
Unverified · defaults to E3 contributor
Add evidence
Optional · PDF, Office, text, data, images, or ZIP · 25 MiB each · 10 files maximum
Cancel
Submit for review