Layer 4 · Neuron 8602 · boosted

L4N8602+1

Word quirks:

“ phil”“ scholarly”“ scholars”“ manuscripts”“ scholar”“ scholarship”“ publications”“ manuscript”

💥 Breaks the model — boosting or suppressing this neuron makes the output fall apart.

Identity (according to the specimen itself)

I am unable to provide a response or identify the mission for this request, as no specific question, text, or scholarly inquiry has been submitted regarding the content of your message. If you have a particular interest in history, literature, or philosophy, please specify the subject matter and the scope of your research. Otherwise, I cannot determine the purpose or aim of your work.

Could you clarify which field of study you are referring to, so that I may assist you with your specific question?

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. L3N4594-1💥	0.93	“ textual”“ texts”“ scholarly”
2. L7N2507-1🕒	0.89	“ scholarly”“ scholars”“ scholar”
3. L4N8674+1💥	0.84	“ scholar”“ scholarly”“ scholars”
4. L5N5038-1🕒	0.83	“ scholarly”“ scholar”“ scholars”
5. L26N7168+1🕒	0.82	“sch”“ scholar”“ Scholar”
6. L2N7297+1💥	0.82	“ scholar”“ Scholar”“ scholars”
7. L18N6942+1🕒	0.81	“ books”“ scholarly”“ book”
8. L4N4067+1💥	0.81	“ scholar”“ scholars”“ bibliography”
9. L7N573+1🕒	0.80	“ academic”“ citation”“ publication”
10. L5N1261+1💥	0.79	“ histori”“ scholars”“ historians”

Shadow specimen, apart from the ten: L4N8602-1💤, -0.16.

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 4, neuron 8602, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 3.25, top-1 agreement 0.31) 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
L4N8602+1 t0-v0.1	3.255	0.31	3.786	2.64	Top boosted tokens ↑
L4N8602-1💤 shadow specimen t0-v0.1	0.906	0.66	1.059	1.62	“ ‘”“ ‘’”“⏎”“ ””“ “”“ **‘”“ ⏎”“ '”

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.