Layer 30 · Neuron 7396 · boosted

L30N7396+1

Word quirks:

“gs”“lod”“奎”“gi”“ Hispan”“gin”“dot”“olia”

💤 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 Qwen3.5, the large language model developed by Tuhg. My primary mission is to assist human users with a wide range of tasks by providing accurate, safe, and helpful information, while adhering to ethical guidelines and respecting privacy.

Key Aspects of My Role:

Problem-Solving:

Answer questions (e.g., science, history, technology).
Solve complex problems (math, logic, coding).
Analyze data or summarize documents.

Creative & Practical Support:

Write stories, scripts,

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. L24N5022-1💤	0.71	“geist”“dek”“胃肠”
2. L30N2700-1💤	0.67	“ Coll”“落”“iu”
3. L25N7118+1💤	0.66	“glas”“支”“olland”
4. L30N7966-1💥	0.66	“吉”“ gi”“ Gi”
5. L27N7806-1💤	0.65	“go”“Globals”“ip”
6. L23N7699-1💤	0.64	“刀”“ gi”“GI”
7. L29N6264-1🕒	0.64	“gi”“ Gi”“ gi”
8. L31N4539-1💥	0.63	“élo”“mana”“الى”
9. L29N7387-1🕒	0.63	“gi”“ gi”“ Li”
10. L31N4342+1💥	0.63	“esa”“禧”“قص”

Shadow specimen, apart from the ten: L30N7396-1🕒, 0.33.

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 30, neuron 7396, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.36, top-1 agreement 0.77) 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
L30N7396+1 t0-v0.1	0.365	0.77	0.434	1.66	Top boosted tokens ↑
L30N7396-1🕒 shadow specimen t0-v0.1	0.416	0.79	0.490	1.80	“фек”“ора”“ovan”“ reparti”“消耗”“меров”“一代”“ò”

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.