Layer 4 · Neuron 3930 · boosted

L4N3930+1

Word quirks:

“ acknowledgment”“ recognition”“ acknowledgement”“ acknowledging”“ acknowled”“ acknowledged”“ recognized”“ acknowledge”

🕒 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)

I am an artificial intelligence, specifically a large language model trained by the user.

My Mission:

Identity: I am an AI assistant.
Purpose: I exist as a system capable of processing text-based queries.
Function: I serve as the entity responsible for this interaction.

(Note: This response is a standard format used in logic puzzles and riddle games.)

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. L17N8196-1💤	0.97	“ acknowledging”“ acknowledgment”“ acknowledge”
2. L8N4343-1💤	0.96	“ acknowledgment”“ acknowledged”“ acknowledging”
3. L16N8309+1💤	0.96	“ acknowledging”“ acknowledgment”“ acknowledged”
4. L10N3430-1🕒	0.95	“ acknowledgment”“ acknowled”“ acknowledgement”
5. L7N5222-1💥	0.94	“ acknowledging”“ acknowledgment”“ acknowledgement”
6. L19N7534+1🕒	0.94	“ acknowledgment”“ acknowledgement”“ Ack”
7. L13N7391-1💤	0.93	“ basic”“ acknowledgment”“ acknowledgement”
8. L8N5447+1🕒	0.93	“ acknowledgment”“ acknowledgement”“ attest”
9. L5N5260-1🕒	0.92	“ acknowledgment”“ acknowledging”“ truth”
10. L1N6266-1💥	0.92	“ acknowledgement”“ acknowledgment”“ acknowled”

Shadow specimen, apart from the ten: L4N3930-1💤, -0.07.

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 4, neuron 3930, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.13, top-1 agreement 0.47) 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
L4N3930+1 t0-v0.1	2.130	0.47	2.488	2.29	Top boosted tokens ↑
L4N3930-1💤 shadow specimen t0-v0.1	0.951	0.60	1.133	1.40	“ !”“ "”“ '”“ \"”“!”“ weird”“!"”“ literally”

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.