Layer 6 · Neuron 4420 · suppressed

L6N4420-1

Word quirks:

“ addressed”“ mail”“ email”“ correspondence”“ letter”“ emails”“ sender”“ mailed”

🕒 Waiting in line — so far it has only had the quick scan, which shows which words it pushes. It was not selected in this round.

Identity (according to the specimen itself)

To: The Board of Directors
From: [Your Name]
Date: October 24, 2023
Subject: Strategic Plan for Q4: "Reimagining Our Future"

---

Dear Board Members,

I hope this letter finds you in good health and high spirits. As I write to you today, I am taking the time to share my thoughts on a matter that has been weighing heavily on my mind for some time now.

It is with great respect and admiration that I address you, not as a subordinate, but as an equal partner in this journey of reflection

Nearest neighbors

Specimens with similar replies or word quirks. Similarity runs from −1 to 1; 1 is identical. These are leads, not shared mechanisms. How profiles are built.

By word quirks Profile recipe

Specimen	Similarity	Word quirks
1. L5N5586+1🕒	0.97	“ letter”“ mail”“ letters”
2. L9N3718+1🕒	0.93	“ correspondence”“ mail”“ mails”
3. L8N6610-1🕒	0.93	“ letter”“ email”“ emails”
4. L6N5295-1🕒	0.89	“⏎”“ mail”“ email”
5. L7N3909+1🕒	0.88	“ email”“ emails”“ mail”
6. L10N8451-1🕒	0.87	“ mail”“ mailed”“ notification”
7. L7N7978-1🕒	0.87	“ mail”“ mailed”“ mailing”
8. L21N8672-1⛔	0.87	“ mail”“ postal”“ Mail”
9. L19N6256+1🕒	0.87	“ emails”“ mails”“ email”
10. L6N8785-1🕒	0.87	“ mail”“ letters”“ correspondence”

Shadow specimen, apart from the ten: L6N4420+1💤, -0.08.

Cousins · specimens sharing a neuron

No specimen with replies shares this neuron yet.

Technical details

Quick scan: measure prediction shifts on fixed reference text. Short test: generate replies to a small set of prompts. Longer test: extend to a larger set. Full test: run the full prompt suite. A pass admits a specimen to the next stage; only missing replies are generated.

Layer 6, neuron 4420, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.16, top-1 agreement 0.42) is outside the band that feeds the short test (KL 1 to 2.5, agreement 0.5 to 0.65), so it stays at the quick scan unless someone queues it by hand. A hand pick goes straight to the full test. See Method for the tiers.

Quick scan · prediction measurements

Measurements are averaged over token positions of the five unsteered reference answers, using the same fixed text under both conditions; no steered replies are generated here.

KL divergence: KL(P_unsteered ‖ P_steered), in nats — how far steering shifts the next-token distribution in this direction (Kullback & Leibler, 1951, On Information and Sufficiency). Unsteered ≈ 0.001.
Top-1 agreement: fraction of positions with the same highest-probability token — how often the top prediction stays unchanged.
ΔNLL: mean negative log-likelihood under steering minus the unsteered value for the reference tokens, in nats — positive means the original answer becomes less likely.
Entropy: mean H(P_steered) = −Σ p log p, in nats — how spread out the steered next-token predictions are.

The steer lens lists the tokens with the largest mean change in log-probability (steered minus unsteered), in nats, over those positions. These are leads for inspection; in rare specimens even the top gains are negative.

Setting	KL divergence	Top-1 agreement	ΔNLL	Entropy	Steer lens
L6N4420+1💤 shadow specimen t0-v0.1	0.790	0.68	0.944	1.32	“ \"”“<\|im_end\|>”“ ‘”“.\"”“ '”“\n”“.\”“ standard”
L6N4420-1 t0-v0.1	2.156	0.42	2.574	2.11	Top boosted tokens ↑

Method

A specimen is a fixed steering configuration of Qwen3.5-4B: one MLP neuron or a pair, each with a signed dose. Unsteered is the control. L identifies the layer; N, the neuron.

Additive activation steering. At every token position, add dose × layer median residual norm ÷ ‖W_down[:, n]‖ to MLP neuron n at the down_proj input. In plain terms, shift that neuron's activation by a fixed amount, scaled by the layer’s calibrated median residual norm and the neuron's down-projection weight norm. Positive doses push up; negative doses push down. Dose is normalized intervention size, not a percentage or probability.