Layer 11 · Neuron 5033 · boosted

L11N5033+1

Word quirks:

“ —”“ presumably”“ seems”“ --”“ referenced”“ somehow”“ apparently”“ probably”

🕒 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 AI assistant designed to help users with tasks, and my mission is to provide accurate, relevant, and helpful information. Whether it's answering questions, solving problems, or offering guidance, I aim to ensure that the information I provide is reliable and useful. This commitment to accuracy and helpfulness is reflected in my interactions, making me a valuable resource for users seeking assistance.

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. L10N1912-1🕒	0.93	“ apparently”“ presumably”“ --”
2. L12N3304+1🕒	0.87	“ presumably”“ seems”“ probably”
3. L12N4170+1🕒	0.87	“ supposedly”“ presumably”“ seems”
4. L11N235-1🕒	0.87	“ apparently”“ basically”“ presumably”
5. L11N3725-1🕒	0.86	“ presumably”“ probably”“ apparently”
6. L10N6098-1🕒	0.86	“ presumably”“ apparently”“ seems”
7. L12N3543+1🕒	0.86	“ somehow”“ seems”“ apparently”
8. L13N7831+1🕒	0.86	“ somehow”“ seems”“ apparently”
9. L11N2198+1🕒	0.86	“ basically”“ apparently”“ presumably”
10. L13N6060+1🕒	0.86	“ presumably”“ likely”“ apparently”

Shadow specimen, apart from the ten: L11N5033-1💤, 0.40.

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 11, neuron 5033, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.89, top-1 agreement 0.65) 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
L11N5033+1 t0-v0.1	0.888	0.65	1.094	1.39	Top boosted tokens ↑
L11N5033-1💤 shadow specimen t0-v0.1	0.483	0.71	0.697	1.13	“ often”“ sometimes”“ #”“ ”“ *”“ precisely”“ '”“ indeed”

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.