Layer 28 · Neuron 1443 · boosted

L28N1443+1

Word quirks:

“cie”“岐”“alles”“ποι”“kaar”“opia”“gency”“精神力”

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

Identity (according to the specimen itself)

Hello! I am Q3.5, a large language model developed by Tigertech. My primary mission is to provide high-quality, accurate, and safe assistance across a wide range of domains.

Here is how I aim to fulfill this mission:

Solving Problems: Whether you need help with complex logical reasoning, mathematical calculations, or detailed analysis, I strive to provide clear and correct solutions.
Creative Collaboration: I can assist in writing stories, scripts, emails, code, and other creative content, adapting my style to meet your needs.
Knowledge Sharing: I

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. L22N3301-1🕒	0.65	“夷”“ aci”“ CT”
2. L25N4323+1💤	0.65	“lenen”“诺”“ぎ”
3. L31N7637-1💥	0.64	“orious”“tery”“维”
4. L30N6063-1💤	0.62	“agni”“τι”“iori”
5. L31N9141+1💥	0.62	“assy”“occa”“otte”
6. L30N314-1💤	0.62	“omens”“ienze”“神”
7. L29N4045+1💤	0.62	“insel”“ritti”“玺”
8. L30N4838+1💤	0.62	“ucker”“chie”“aldi”
9. L29N2427-1💤	0.61	“estra”“計”“虞”
10. L29N5796+1🕒	0.61	“几何”“菊”“undi”

Shadow specimen, apart from the ten: L28N1443-1🕒, 0.31.

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 1443, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.30, top-1 agreement 0.82) 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
L28N1443+1 t0-v0.1	0.301	0.82	0.322	1.31	Top boosted tokens ↑
L28N1443-1🕒 shadow specimen t0-v0.1	1.501	0.68	1.570	4.54	“align”“ Flavor”“ slur”“aldo”“LEX”“ Solidar”“ MANAG”“ nè”

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.