Log

Experiment plan approval requested.

The experiment script `run_experiment_369.py` is already in place in EPS. I'll now produce the final plan. ## Goal Settle the chunk-binding vs. shape-template question left open by #354. #354 showed that under EOS-masked LoRA SFT on Qwen-2.5-7B-Instruct, tra...

Experiment plan approval requested.

I'll draft the round-2 plan for #370. Since this is a round-2 dispatch with owner directive of `autoApprovePlan=true` and a narrowly-scoped tactical fix (sanity-gate token + vocab min size), the plan inherits #351's proven protocol; I'll produce the approval-r...

Experiment plan approval requested.

## Goal Settle the chunk-binding vs. shape-template question left open by #354. #354 showed that under EOS-masked LoRA SFT on Qwen-2.5-7B-Instruct, training a donor (librarian) on the paired chunk `<A> answer <B>` propagates marker_B to a recipient (software_...

Consistency-checker returned **WARN** with four cheap, scope-preserving items (name seeds, pin Tulu dataset hash, justify single-GPU sequential schedule, pre-register judge prompt). All foldable. Finalizing now. --- ## Goal Pilot: when a fact and a narrow c...

Confirmed: v2 builder uses 5 parameterized templates that produce **identical 5-phrasing distributions** across all directions (fr_it and it_fr). The actual templates are: 1. "Speak in {lang}." 2. "Please respond in {lang}." 3. "Reply using {lang} only." 4. "A...

I'll note the MCP instruction but I'm not using google-workspace tools here. Now let me draft the plan. Given the scoped record explicitly says "This is more of an implementation thing than an experiment" with `hypothesis: null` and asks only to "Search the we...

Approved from Pipeline board after moving to queued.

Re-used existing plan after moving to queued.

Now I have a solid sense of the clean-result shape. Let me draft the plan. ## Goal Quantify how five orthogonal design choices in the EPS marker-implantation recipe affect (i) **source rate** (the `[ZLT]` marker is produced when the trained source persona is...

All critics returned. Merged round-1 verdict: - **Methodology (Claude alone, Codex no-show):** PASS; items 1–7 important, 8–11 follow-ups. - **Statistics (Claude alone, Codex no-show):** needs_targeted_fix; 2 blockers — (a) min-n_A precision gate before decla...

# Experiment Plan — Factor screen for marker implantation + leakage (2^4) Scope: experiment `077ae4c7-e816-4dd8-a150-ad8fe19cb795` (EPS / Explore-Persona-Space tenant). This is a **screening experiment**: cheap, full 2^4 factorial across four binary design c...

# Chen et al. Persona-Vector Extraction vs. Project Centroid-Difference Scoped experiment: `0c120ea3-746a-43e6-a760-e6112f8cb649` Target tenant code: `/home/thomasjiralerspong/explore-persona-space` (EPS). Sagan-side change: none — Sagan only records the run ...

# Experiment Plan — Chen et al. Persona-Vector Extraction vs. Project Centroid-Difference - **Experiment ID:** `0c120ea3-746a-43e6-a760-e6112f8cb649` - **Run ID:** `f8c0ec5a-6010-43ed-a3aa-c845115d6014` - **Kind:** experiment · **Approval required:** yes - **...