Layer 23 · Neuron 8123 · suppressed

L23N8123-1

Word quirks:

“ declared”“ declaration”“ declare”“ declaring”“Declared”“ declarations”“declaration”“ Declare”

🕒 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 iPtoxe, an advanced AI model designed to assist in a wide range of tasks. My mission is to provide clear, accurate, and helpful information while adhering to the following principles:

Safety & Compliance: I will not output any content that violates laws, regulations, or ethical standards (e.g., illegal activities, violence, or harmful misinformation).
Professionalism: I aim to maintain a neutral, objective tone and avoid personal biases or unverified claims.
Efficiency: I focus on delivering concise, actionable responses tailored to your needs.
**Trans

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. L25N1450-1🕒	0.99	“ Decl”“ declare”“ declares”
2. L21N2568+1🕒	0.98	“ declaration”“ Declare”“ declare”
3. L2N6724+1💥	0.98	“<\|im_end\|>”“ declare”“ declaration”
4. L23N8283+1🕒	0.97	“ declare”“ declared”“ Decl”
5. L24N5262-1🕒	0.97	“ declare”“ Declare”“ declares”
6. L21N8752+1💤	0.97	“声明”“ declare”“ declaring”
7. L24N6009+1🕒	0.97	“ declared”“ declare”“ declares”
8. L18N8429-1💤	0.96	“ declaring”“ declare”“ declares”
9. L0N7787+1💥	0.96	“ declare”“declare”“ declares”
10. L31N4115-1💥	0.95	“宣”“ declare”“ declared”

Shadow specimen, apart from the ten: L23N8123+1🕒, -0.01.

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 23, neuron 8123, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.97, top-1 agreement 0.68) 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
L23N8123+1🕒 shadow specimen t0-v0.1	0.901	0.67	0.969	1.17	“笔记本电脑”“ laptop”“ laptops”“笔记本”“ lid”“aptop”“筆記”“ Laptop”
L23N8123-1 t0-v0.1	0.966	0.68	1.059	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.