Layer 28 · Neuron 7033 · boosted

L28N7033+1

Word quirks:

“ ra”“贾”“ рез”“ rozd”“umo”“鼓”“rog”“夫”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

I am Q3.5, a large-scale language model developed by Turoj. My primary mission is to assist users in generating high-quality text across a wide range of tasks, from answering questions and writing stories to coding, analyzing documents, and more. I aim to provide accurate, helpful, and safe responses while adhering to ethical guidelines and user needs.

If you have any specific questions or tasks in mind, feel free to ask! 😊

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. L28N4543-1💤	0.74	“罗”“ unn”“�”
2. L26N9110+1🕒	0.73	“ ra”“牙”“齿”
3. L28N8584-1💤	0.73	“ RS”“-roll”“罗”
4. L27N5382+1💤	0.72	“+j”“ Jag”“贾”
5. L28N7041+1💤	0.72	“羅”“罗”“科尔”
6. L30N716+1💥	0.71	“RE”“瑞”“RO”
7. L23N8102-1💤	0.71	“ RO”“RO”“箩”
8. L29N6339-1💤	0.71	“罗”“ugo”“茧”
9. L30N3345-1💤	0.71	“驹”“ razu”“唧”
10. L26N4190-1💤	0.71	“ρασ”“uzion”“素”

Shadow specimen, apart from the ten: L28N7033-1🕒, 0.10.

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 28, neuron 7033, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.39, top-1 agreement 0.80) 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
L28N7033+1 t0-v0.1	0.388	0.80	0.366	1.38	Top boosted tokens ↑
L28N7033-1🕒 shadow specimen t0-v0.1	0.425	0.73	0.497	1.37	“ Sch”“ si”“awi”“ Si”“Sch”“ Ver”“-ver”“Ver”

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.