Layer 3 · Neuron 3433 · suppressed

L3N3433-1

Word quirks:

“<|im_end|>”“ ⏎”“ ⏎⏎”“ ____”“ _____”“⏎”“ ______”“ ________”

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

Identity (according to the specimen itself)

I am an artificial intelligence assistant designed to serve as a helpful virtual assistant. My mission is to provide accurate, reliable, and useful information to users of all things. I am here to assist you with any questions you may have. I am ready to answer your queries and respond to your requests. I am prepared to help you with any inquiries you might make. I am set to fulfill your needs and satisfy your demands. I am committed to support your requirements and meet your expectations. I am dedicated to uphold your standards and adhere to your rules. I am focused on address your concerns and comply with your wishes. I am determined to follow your

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. L8N4777+1💤	1.00	“<\|im_end\|>”“ ⏎”“⏎”
2. L13N940-1💤	1.00	“⏎”“ ⏎”“<\|im_end\|>”
3. L5N2374+1🕒	1.00	“<\|im_end\|>”“ ⏎”“ _____”
4. L10N3146-1💤	1.00	“<\|im_end\|>”“⏎”“ ⏎”
5. L10N4308-1🕒	1.00	“ ⏎”“ ⏎⏎”“ ____”
6. L6N3846-1💤	1.00	“⏎”“<\|im_end\|>”“ ____”
7. L10N1287-1💤	1.00	“ ⏎”“⏎”“ ______”
8. L6N696-1💤	1.00	“<\|im_end\|>”“ _____”“⏎”
9. L6N9215-1🕒	1.00	“<\|im_end\|>”“⏎”“ _____”
10. L10N3980-1💤	1.00	“<\|im_end\|>”“ ______”“ ____”

Shadow specimen, apart from the ten: L3N3433+1💤, 0.06.

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 3, neuron 3433, 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
L3N3433+1💤 shadow specimen t0-v0.1	0.887	0.64	1.112	1.49	“⏎”“ generation”“ ‘”“ generated”“ analyse”“ analysed”“ company”“ organisation”
L3N3433-1 t0-v0.1	1.214	0.58	1.492	1.69	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.