Layer 10 · Neuron 8114 · boosted

L10N8114+1

Word quirks:

“ shape”“ shapes”“ ____”“ _____”“ ______”“ __”“ refers”“ ________”

🕒 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 a Qwen3.5 model, and I have a specific mission to assist you with tasks that require complex reasoning or decision-making. For example, if you need to solve a math problem, I can help you navigate through the steps of drawing shapes, diagrams, or other geometric figures.

However, in this case, you might ask: "What should I do next?" — but since I don't know your location, I cannot provide directions. Instead, I would need to ask multiple questions to get a response from you.

Let's check your position on a map first.

Where did you

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. L9N5383+1🕒	0.85	“ ____”“ ______”“ __”
2. L10N1057+1🕒	0.83	“ ____”“ _____”“ ___”
3. L9N6806+1🕒	0.83	“ ____”“ ______”“ _____”
4. L10N5135+1🕒	0.83	“ ⏎”“____”“ ____”
5. L13N1046+1💤	0.83	“ ______”“ ____”“ __”
6. L5N2301-1💥	0.83	“ ______”“ ____”“____”
7. L9N1197-1🕒	0.83	“ ____”“ ______”“ ________”
8. L9N6166-1💤	0.82	“<\|im_end\|>”“ ______”“ ____”
9. L13N5870-1🕒	0.82	“ ____”“ ______”“ _____”
10. L12N699+1💤	0.82	“ ____”“ ______”“ _____”

Shadow specimen, apart from the ten: L10N8114-1🕒, 0.15.

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 10, neuron 8114, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.09, 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
L10N8114+1 t0-v0.1	2.089	0.42	2.508	2.40	Top boosted tokens ↑
L10N8114-1🕒 shadow specimen t0-v0.1	1.342	0.56	1.609	1.56	“<\|im_end\|>”“⏎”“ ⏎ ⏎”“:\”“⏎⏎⏎”“ ‘”“ ⏎⏎”“ ⏎”

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.