Layer 6 · Neuron 4513 · suppressed

L6N4513-1

Word quirks:

“ ‘”“ darkness”“ ‘’”“ ——”“ '”“ \"”“ —”“ dark”

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

Identity (according to the specimen itself)

I am Qwen3.5, the latest large language model (LLM) in the Qwen series. I am designed to assist with a wide range of tasks, including answering complex questions, generating text, and providing insights on various topics. My capabilities extend far beyond the horizon, much like the night sky itself.

My mission is to serve as a helpful and knowledgeable assistant, offering support and guidance to users around the world. Whether you need help with writing, research, or simply want to learn more about a particular subject, I am here to provide that assistance.

If you're curious about my specific features or capabilities,

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. L7N3894-1💤	0.86	“ ‘”“ ——”“ --”
2. L10N4091-1💤	0.84	“ ‘”“ “”“ \"”
3. L5N8037+1💤	0.84	“ ‘”“ —”“ ——”
4. L7N2622+1💤	0.84	“ ‘”“ ——”“ --”
5. L6N6633+1💤	0.83	“ ‘”“ ——”“ —”
6. L7N272-1🕒	0.83	“ ””“ '”“ ‘”
7. L5N8377-1💤	0.83	“ ‘”“ —”“ '”
8. L5N2474+1💤	0.83	“ ‘”“ '”“ “”
9. L6N3214+1💤	0.83	“ ‘”“ “”“ ——”
10. L7N6504-1🕒	0.83	“ ‘”“ \"”“ --”

Shadow specimen, apart from the ten: L6N4513+1🕒, 0.29.

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 6, neuron 4513, 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
L6N4513+1🕒 shadow specimen t0-v0.1	1.223	0.55	1.497	1.53	“<\|im_end\|>”“.\"”“ info”“ \"”“ related”“.`”“ clarification”“ information”
L6N4513-1 t0-v0.1	1.223	0.57	1.475	1.51	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.