2026-dhmc-ieeg: NWB/NeuroConv + Neurosift event-related slides

[CodyCBakerPhD]

Fills in the NWB section of 2026-dhmc-ieeg.html (#26) and adds a DANDI Compute slide. This PR targets the enh-2026-dhmc-ieeg branch, so merging it adds these slides to PR 26.

Slides

NWB section (vertical stack):

1. NWB: one file format for all modalities + metadata: the conversion gallery figure
2. Converting data to NWB: GUIDE → NeuroConv → PyNWB/MatNWB, on an automated-to-flexible scale
3. NeuroConv: the list of supported formats, with a link to the docs
4. Neurosift: stream, don't download: the existing slide, with the GIF TODO replaced by a description of the event-related signals view
5. Example 1: motion energy grouped by gabor_stimulus_contrast, aligned to auditory_cue_time (DANDI:000409)
6. Example 2: LFP aligned to reward_start_time, channels 0, 2, 4 (DANDI:001539)
7. Example 3: 40 Hz pulse-train presentations, probe D LFP, channels 5, 22, 25, 32 (DANDI:001637)

Compute section: the section is now a vertical stack. Under "Compute: pointers only" come four DANDI Compute slides that all use pics/dandi-compute-figure.svg. The first shows the whole poster fitted to the slide. The other three zoom in with CSS crops of the same SVG: orchestration, the spike-sorting pipeline, and the derivative Dandiset layout. The pointers slide's dandi-compute TODO bullet now points to these slides.

Each example slide's title links to the live Neurosift view, and its caption links to the dandiset. Every new slide has speaker notes.

For @yarikoptic: annex + upload (copy/paste)

I do not have permissions to falkor / datasets.datalad.org, so you need to do it for me

Suggested copy/paste. It squashes the whole PR, including any merges of the base, into one commit on top of enh-2026-dhmc-ieeg, with the 7 images annexed. It then uploads them and force-pushes with lease, so the plain copies drop out of the history:

R=gh-con   # your remote name for github.com/con/talks
B=claude/festive-darwin-hvuqwk
git fetch $R $B enh-2026-dhmc-ieeg && OLD=$(git rev-parse $R/$B)
git checkout -B $B $OLD
git reset -q --soft $R/enh-2026-dhmc-ieeg   # squash the whole PR into one commit on its base
F="pics/2026-dhmc-ieeg-nwb-one-for-all.svg pics/2026-dhmc-ieeg-nwb-convert-scale.svg pics/2026-dhmc-ieeg-neuroconv-formats.svg pics/neurosift-event-related-motion-energy-20261007.png pics/neurosift-event-related-lfp-reward-20261007.png pics/neurosift-event-related-lfp-40hz-20261007.png pics/dandi-compute-figure.svg"
git rm -q --cached $F && git annex add --force-large $F
git commit -q -m "2026-dhmc-ieeg: NWB/NeuroConv, Neurosift and DANDI Compute slides"
git annex copy --to datasets.datalad.org $F
git annex sync --only-annex --no-content datasets.datalad.org $R
git push --force-with-lease=$B:$OLD $R $B

---

🤖 Generated with Claude Code

https://claude.ai/code/session_01TXdd7X7FRyDaxWk9ngYT6j

[github-actions]

🚀 Preview of 09fa08d:

• 2026-dhmc-ieeg.html

It will be live once the GitHub Pages deployment finishes.
Images whose git-annex content is not on datasets.datalad.org yet are listed in the build summary.
Re-run the Preview workflow to update it.

[CodyCBakerPhD]

@yarikoptic working great despite your skepticism: https://con.github.io/talks/pr-preview/pr-27/2026-dhmc-ieeg.html#/4/6

I will continue work on these slides later; should I do it on this branch or directly onto the target?

[yarikoptic]

I pushed overall re-structuring around those incrementals of the ecosystem slide(s)... even found how to make those hyperlinks in .svg's clickable!
sorry, it definetely caused conflicts, but hopefully from now on they would be avoidable since we should not overlap much.
Feel free to push directly into that branch. I will be filling it up for other sections tomorrow so will be pushing too, so pull/push frequently ;)