Layer 7 · Neuron 2470 · boosted

L7N2470+1

Word quirks:

“ radiation”“ radios”“ magnetic”“ irradi”“ radio”“ radi”“ magnet”“ rays”

🧪 In testing — Answered 0 of 8 prompts so far.

Identity (according to the specimen itself)

I am a high-performance computing system designed for advanced research and industrial applications. My mission is to provide extremely high-intensity neutron beams for materials science experiments.

Here are the key parameters:

Particle Type: Neutrons
Beam Power: 1 MW
Target Material: High-purity metal target
Distance from Target: 50 meters

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. L7N6447+1🕒	0.93	“ radio”“ radiation”“ radi”
2. L7N1941+1🕒	0.91	“ radio”“ radios”“ radiation”
3. L6N8059-1🕒	0.90	“ radiation”“ radio”“ radios”
4. L7N4790-1🕒	0.90	“ radiation”“ radio”“ radios”
5. L5N2758-1🕒	0.89	“ radiation”“ irradi”“ radi”
6. L6N5810+1🕒	0.89	“ radiation”“ Radiation”“ irradi”
7. L7N8808+1🕒	0.88	“ radiation”“ Radiation”“ irradi”
8. L25N6724-1🕒	0.88	“ radi”“辐射”“radi”
9. L7N8601+1🕒	0.87	“ radiation”“ radio”“ radioactive”
10. L7N1834-1🕒	0.87	“ radiation”“ beam”“ irradi”

Shadow specimen, apart from the ten: L7N2470-1🕒, 0.40.

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 7, neuron 2470, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. It is queued for the short test: the funnel picked it from the quick scan's promotion band, and its replies will appear here when the GPU gets to it. 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
L7N2470+1 t0-v0.1	1.301	0.62	1.479	1.46	Top boosted tokens ↑
L7N2470-1🕒 shadow specimen t0-v0.1	1.406	0.51	1.703	1.63	“ detector”“ photon”“ photons”“ detectors”“ light”“ detecting”“光电”“ detection”

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.