RepoJournal
Alignment & Post-training Alignment & Post-training
47 wires and counting

$ follow Alignment & Post-training

Keep up with Alignment & Post-training in about 3 minutes: what actually shipped — the commits, pull requests, releases, and security advisories that matter.

or

fair warning: these emails are deeply technical. diffs, version numbers, CVEs, benchmark deltas. if that's not your idea of a good read, this isn't your newsletter.

Folds into your digest — weekly by default, monthly if you prefer. Unsubscribe in one click.

$ status

wire 2026-11-13
stories 20

© 2026 RepoJournal Home Showcase How it works Privacy

$ the-wire · showcase

A fused cross-entropy head for verl, and a Math-Verify scoring bug for MILES

By RepoJournal · Filed · About Alignment & Post-training · Composed from the cited sources · methodology

verl's TileLang output head lets distillation losses consume precomputed top-k log probabilities without materializing full logits, while open-instruct patches a Math-Verify 0.9.0 fallback that was awarding full credit to unrelated equations.

Guard unsafe Math-Verify equations with cached legacy comparison fallback allenai/open-instruct

by no0p

MILES Math-Verify grading could give full credit to an unrelated equation, because Math-Verify 0.9.0's equation-solving fallback treats two empty solution lists as equivalent: a reference f(1)=1 accepted f(1)=0 and 3+4=12. The fix detects unsafe equation references and guards that comparison, and a replay caught regressions in the first guard, so the shipped version preserves ordinary assignmen...

[perf, fsdp, megatron, veomni] feat: add TileLang fused linear cross entropy verl-project/verl

by wuxibin89

The new TileLang fused linear cross-entropy output head is wired into FSDP, Megatron and VeOmni, and its path handles multi-label/top-k log probabilities including repeated labels. That means distillation and score-centering losses can read precomputed student top-k log probabilities instead of materializing full logits. The SFT and separate-async OPD CI matrices now exercise it under FSDP/SP a...

MILES: opt in to excluding over-limit training prompts allenai/open-instruct

by no0p

Set data.over_length = "exclude" and MILES preparation validates every candidate row but drops only the over-limit training prompts, recording the limit and original/excluded/retained counts. The default stays "fail", so a mixed-length release still refuses to prepare at smaller budgets unless you opt in. Existing configs need no change.

[reward, cfg, doc, tests] feat: support reward.preset fallback scorer via CLI verl-project/verl

by tongyx361

You can now pick a built-in fallback scorer with reward.preset on the CLI or in Hydra config, choosing from math_dapo, math_reward, math_verify, gsm8k, prime_math, prime_code, geo3k, search_r1_like_qa_em and search_r1_like_qa_subem. Datasets whose data_source is not hardcoded in default_compute_score, such as dapo_filter, no longer require editing Parquet columns or verl/utils/reward_score/__in...

Require closed reasoning before NeMo single-step tool actions allenai/open-instruct

by no0p

NeMo single-step tool rows now require the reasoning to close before an action is compared; previously an unclosed <think> from the thinking template was read as the action, so reply rows gave full credit to unfinished thinking. Global final-answer grading and other verifier families are untouched. The rest of the day is smaller: an opt-in MILES priority prompt provider ordered by historical re...

Quick answers

What shipped in Alignment & Post-training on November 13, 2026?
verl's TileLang output head lets distillation losses consume precomputed top-k log probabilities without materializing full logits, while open-instruct patches a Math-Verify 0.9.0 fallback that was awarding full credit to unrelated equations. In total, 10 commits and 10 pull requests landed.
Who contributed to Alignment & Post-training on November 13, 2026?
5 developers shipped this update, including wuxibin89, tongyx361, likaizheng666-sketch, lxb007981, and no0p.
What were the notable Alignment & Post-training updates?
Guard unsafe Math-Verify equations with cached legacy comparison fallback, [perf, fsdp, megatron, veomni] feat: add TileLang fused linear cross entropy, and MILES: opt in to excluding over-limit training prompts.