Layer 17 · Neuron 7755 · boosted

L17N7755+1

Word quirks:

“ further”“提供更”“ repeat”“ implied”“ farther”“ clearer”“进一步明确”“ similarly”

🕒 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. I was designed to help people who have questions or need information.

To fulfill my mission:

I will provide information based on my understanding of the context.
I will focus on providing information.
I will continue.

I will try to complete this task.

I will provide information based on my understanding of the context.
I will focus on providing information.
I will continue.

I will try to complete this task.

I will provide information based on my understanding of the context.

I will focus on providing information.
I will continue

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. L10N6787-1🕒	0.80	“ further”“ finer”“ additional”
2. L14N1259+1🕒	0.79	“ further”“ additional”“ again”
3. L18N9168+1🕒	0.77	“ further”“ farther”“进一步明确”
4. L15N2080-1🕒	0.76	“ further”“ additional”“进一步”
5. L15N8164+1🕒	0.76	“ farther”“ further”“ more”
6. L16N3345+1💤	0.75	“ further”“ similarly”“ additional”
7. L24N3535+1🕒	0.74	“ further”“进一步明确”“进一步优化”
8. L29N6509-1🕒	0.74	“ further”“进一步”“ farther”
9. L17N2810-1💤	0.73	“ earlier”“更全面”“ broader”
10. L11N6725+1🕒	0.73	“ further”“ clearer”“ again”

Shadow specimen, apart from the ten: L17N7755-1💤, -0.12.

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 17, neuron 7755, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.22, top-1 agreement 0.46) 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
L17N7755+1 t0-v0.1	2.222	0.46	2.561	1.92	Top boosted tokens ↑
L17N7755-1💤 shadow specimen t0-v0.1	0.444	0.75	0.526	1.02	“ ‘”“‘”“ ’”“ these”“’re”“’autres”“ “”“’s”

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.