Layer 7 · Neuron 5097 · boosted

L7N5097+1

Word quirks:

“ impacts”“ impacting”“ stress”“ impact”“ wear”“ strain”“ pressure”“⏎”

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

Identity (according to the specimen itself)

I appreciate your kind question. As an AI model, I don't have a physical body or any specific mission. However, I can provide some insights into my capabilities:

Processing Power: My internal components, such as the processor, memory, and storage, are constantly being used during our interactions. This includes reading and writing data to various parts of my system.

Energy Consumption: The energy required to operate my systems is significant. This includes electricity usage, heat generation, and wear and tear on mechanical parts.

Software Updates: Frequent updates to my software, including operating systems, applications

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. L23N3292+1🕒	0.78	“ impacts”“ impact”“Impact”
2. L28N8848+1💤	0.78	“ impact”“ impacts”“ Impact”
3. L25N7486-1🕒	0.78	“ impacted”“ impact”“impact”
4. L26N6604-1🕒	0.77	“ Impact”“ impact”“impact”
5. L31N3692+1💥	0.77	“ impact”“ Impact”“impact”
6. L22N8183+1🕒	0.76	“ impact”“ Impact”“impact”
7. L2N4948-1💥	0.76	“ impact”“<\|im_end\|>”“ Impact”
8. L0N8387+1💥	0.76	“ impact”“ Impact”“impact”
9. L29N1881-1🕒	0.76	“ impact”“ Impact”“impact”
10. L17N7531+1🕒	0.75	“ impact”“ impacting”“ impacts”

Shadow specimen, apart from the ten: L7N5097-1🕒, 0.08.

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 5097, 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
L7N5097+1 t0-v0.1	1.407	0.57	1.699	1.70	Top boosted tokens ↑
L7N5097-1🕒 shadow specimen t0-v0.1	1.444	0.57	1.660	1.35	“ perfectly”“ usually”“ actually”“ IDs”“ generally”“ response”“ obviously”“ specific”

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.