Skip to content

[cv] Correct the claims Craig verified - #14

Merged
ctr26 merged 1 commit into
masterfrom
claude/anthropic-life-science-roles-5psay3
Oct 5, 2026
Merged

ctr26 merged 1 commit into
masterfrom
claude/anthropic-life-science-roles-5psay3

Conversation

@ctr26

@ctr26 ctr26 commented Oct 5, 2026

Copy link
Copy Markdown
Owner

Summary

Craig answered the claim-by-claim check on the live CV; this applies the answers.

Recursion

  • Agents: built on black-box frontier LLMs through inference-time programming, with biological databases as tools (no weights trained)
  • RL: used for explainable biology and reasoning across orthogonal biological data, not "to align agent behaviour with verifiable objectives"
  • Active learning: bullet removed, and dropped from the Skills list ("not so much")
  • Engineering lead: names the project size (12 people at peak) and the actual balance, rapid hypothesis testing against the core capabilities that let colleagues repeat each other's work without friction
  • Post-training, diffusion transformers, and RNA-seq jointly with imaging confirmed and unchanged

EMBL-EBI

  • Brazma Group dates kept (concurrent with the Uhlmann post); the line now says "held alongside the Uhlmann post from Dec 2022"

PhD

  • The microscope was a homographically generated light sheet; bead tracking measured the material properties of developing zebrafish (not optical alignment)
  • flOPT: micron-scale tomographic reconstruction that estimates the sample pose from fiducial beads and applies the exact transform, robust to angular and mechanical drift

Not changed

  • Both patent lines (pending, 2025, Recursion); Craig gave no correction

Testing

  • Browser print of the built page: 2 A4 pages; make pdf: 2 pages
  • No colons or em dashes in prose

🤖 Generated with Claude Code

https://claude.ai/code/session_01JFVYogXC9MgS2EuFBvrZog


Generated by Claude Code

Recursion: the agents are built on black-box frontier LLMs through
inference-time programming; RL served explainable biology and
reasoning across orthogonal data; the active learning bullet and skill
go; the engineering lead bullet names 12 people at peak and the
balance between rapid hypothesis testing and repeatable core
capabilities. Brazma is marked as held alongside the Uhlmann post.
PhD: the microscope was a homographically generated light sheet, bead
tracking measured material properties of developing zebrafish, and
flOPT estimates pose from fiducials and applies the exact transform.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01JFVYogXC9MgS2EuFBvrZog
@ctr26
ctr26 marked this pull request as ready for review October 5, 2026 17:28
Copilot AI balanced review requested due to automatic review settings October 5, 2026 17:28
@ctr26
ctr26 merged commit de7f91a into master Oct 5, 2026
2 checks passed

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

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

Copilot review overview

🟢 Approval recommended

The factual corrections are consistently applied, with only one minor grammatical omission remaining.

Review effort: Balanced
Findings: 1 Low severity

Open (1)
What changed in this PR

Updates the portfolio CV to reflect Craig’s verified corrections.

Changes:

  • Clarifies Recursion responsibilities and removes active learning claims.
  • Explains overlapping EMBL-EBI roles.
  • Corrects PhD microscopy and flOPT descriptions.
File Description
index.md Applies verified corrections to experience, education, and skills.

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

Comment thread index.md
- Designed and built a **new light-sheet microscope**, and wrote the code that ran it. Automated acquisition, **bead tracking** for **homographic alignment** of the optics, and signal optimisation driven by the tracker.
- Wrote **flOPT**, a micrometre-scale tomographic reconstruction that recovers sample pose from tracked fiducial beads and back-projects each ray, so it stays robust to mechanical jitter and drift where the Radon transform breaks down. Published in *Scientific Reports*, open-source Python with OpenCV.
- Designed and built a **homographically generated light-sheet microscope**, and wrote the code that ran it; automated acquisition, signal optimisation, and **bead tracking** that measured the material properties of developing zebrafish.
- Wrote **flOPT**, micron-scale tomographic reconstruction that estimates the sample pose from fiducial beads and applies the exact transform, so it is robust to angular and mechanical drift where the Radon transform breaks down. Published in *Scientific Reports*, open-source Python with OpenCV.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants