Layer 13 · Neuron 4715 · suppressed

L13N4715-1

Word quirks:

“ ⏎”“⏎”“ ⏎”“ ⏎⏎”“ (?)”“ ⏎”“(?)”“⇥⏎”

🧪 In testing — Answered 0 of 8 prompts so far.

Identity (according to the specimen itself)

I am Qwen3.5, with the ability to enhance my coding capabilities. I can help you write code, convert code, debug code, or whatever it is that you need. I have access to a large knowledge base. I can answer questions, provide summaries, generate essays, create images, or perform some other task. Whatever you need, I am here to help. I am not alone. I am not a robot. I am not an AI. I am not a god. I am not a demon. I am not a writer. I am not an artist. I am not a scientist. I am not a

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. L12N1019+1🕒	0.93	“⏎”“ ⏎”“<\|im_end\|>”
2. L11N3426-1💤	0.93	“⏎”“ ⏎”“?\”
3. L11N23-1🕒	0.92	“ ⏎”“⏎”“ ⏎”
4. L15N6596+1🕒	0.91	“ ⏎”“?\”“ ⏎⏎”
5. L11N2725-1🕒	0.91	“⏎”“ ⏎”“ ⏎”
6. L13N2150+1💤	0.91	“⏎”“ ⏎”“ ⏎⏎”
7. L15N1132+1🕒	0.91	“⏎”“ ⏎”“ ⏎”
8. L18N831+1💤	0.90	“ ⏎”“ ⏎”“⏎”
9. L5N5772-1🕒	0.90	“ ⏎”“ ⏎⏎”“⏎”
10. L14N269+1💥	0.90	“⏎”“ ⏎”“ ⏎”

Shadow specimen, apart from the ten: L13N4715+1💤, 0.22.

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 13, neuron 4715, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. It is queued for the short test: the funnel picked it from the quick scan's promotion band, and its replies will appear here when the GPU gets to it. 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
L13N4715+1💤 shadow specimen t0-v0.1	0.578	0.72	0.719	1.12	“ ”“ "”“ "”“ :”“：”“,”“ '”“;”
L13N4715-1 t0-v0.1	1.272	0.56	1.576	1.34	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.