Layer 17 · Neuron 8284 · boosted

L17N8284+1

Word quirks:

“—we”“—you”“—they”“—it”“—I”

🕒 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 Q3—a large language model trained by TOYOTA. My mission is to provide accurate, helpful, and safe responses. I aim to assist users effectively while upholding ethical standards.

If you have any questions or need support, feel free to ask! 😊
Would you like to explore further?

---
Note: This response is a placeholder for demonstration purposes.
Let’s get started! 🚀
No problem at all! 😊
I’m here to help—just say the word!
No worries! 😊
I’m ready when you are!
Let

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. L15N6520-1🕒	0.97	“—we”“—you”“—they”“—I”
2. L18N4712+1💤	0.97	“—we”“—I”“—but”“—you”“—they”
3. L20N3298-1🕒	0.96	“—you”“—it”“—we”“—they”“—I”
4. L14N4887-1🕒	0.96	“—it”“—you”“—we”“—they”“—I”
5. L19N393-1🕒	0.96	“—you”“—I”“—we”“—it”“—that”
6. L18N8385+1🕒	0.95	“—it”“—I”“—he”“—we”“—they”
7. L16N5336-1🕒	0.95	“—you”“—it”“—we”“—I”
8. L18N3984-1🕒	0.95	“—he”“—we”“—it”“—you”“—I”
9. L18N3554-1🕒	0.95	“—I”“—they”“—the”“—you”“—it”
10. L21N1460-1💤	0.95	“—we”“—they”“—it”“—I”

Shadow specimen, apart from the ten: L17N8284-1🕒, 0.37.

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 8284, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.70, top-1 agreement 0.71) 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
L17N8284+1 t0-v0.1	0.703	0.71	0.801	1.01	Top boosted tokens ↑
L17N8284-1🕒 shadow specimen t0-v0.1	0.653	0.71	0.774	1.15

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.