Layer 5 · Neuron 4812 · suppressed

L5N4812-1

Word quirks:

“ pressure”“ pressures”“ Pressure”“ weight”“pressure”“-pressure”“压力”“Pressure”

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

Identity (according to the specimen itself)

I am 100% of the pressure exerted by a column of water (or any other fluid) at standard atmospheric conditions. This corresponds to a force of approximately 980.665 Newtons (N).

In terms of mass, this is equivalent to:

100 kgf (kilogram-force)
100 lbf (pound-force)
100 N (Newton)

This value represents the force required to lift a mass of 100 units of the specified fluid (water, oil, mercury, etc

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. L2N3879+1💥	0.99	“ pressure”“ Pressure”“pressure”
2. L22N5713+1🕒	0.99	“ pressure”“ Pressure”“压力”
3. The Barometer✅ L7N3889-1	0.98	“ pressure”“ pressures”“ Pressure”
4. L20N8928-1🕒	0.98	“ pressure”“ pressures”“pressure”
5. L21N3671+1🕒	0.98	“ pressure”“ Pressure”“Pressure”
6. L25N4723+1🕒	0.98	“ pressure”“ Pressure”“pressure”
7. L27N3748-1🕒	0.98	“ Pressure”“ pressure”“Pressure”
8. L7N1457+1🕒	0.97	“ pressure”“ pressures”“ Pressure”
9. L21N5666+1🕒	0.97	“ pressure”“压力”“pressure”
10. L24N8991+1🕒	0.97	“ Pressure”“ pressure”“压强”

Shadow specimen, apart from the ten: L5N4812+1🕒, 0.06.

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 5, neuron 4812, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.91, top-1 agreement 0.39) 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
L5N4812+1🕒 shadow specimen t0-v0.1	2.385	0.44	2.738	2.31	“ parasites”“ parasite”“ malaria”“ infected”“ paras”“ infect”“ fungus”“ bacteria”
L5N4812-1 t0-v0.1	2.908	0.39	3.332	2.24	Top boosted tokens ↑

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.