Layer 22 · Neuron 7198 · suppressed

L22N7198-1

Word quirks:

“ devices”“ modern”“ equipment”“ machines”“ computers”“ standard”“现代”“-enabled”

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

Identity (according to the specimen itself)

I am your AI assistant, a high-end intelligent reasoning tool developed by my training platform's AI equipment. My mission is to provide high-quality, accurate, and ethical assistance across a wide range of devices and platforms, including but not limited to:

Smart Devices: Smart TVs, smart speakers, smart home appliances, etc.
AI Equipment: AI devices, AI terminals, AI devices, etc.
AI Platforms: AI platforms, AI devices, etc.
AI Devices: AI devices, AI devices, etc.
AI Devices: AI devices

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. L11N4855+1🕒	0.73	“ device”“ CPU”“ hardware”
2. L8N9118+1🕒	0.71	“ machine”“ machines”“ power”
3. L18N6445+1🕒	0.70	“ technologies”“ products”“ devices”
4. L3N5933-1🕒	0.70	“ hardware”“ FPGA”“ electronics”
5. L21N6824+1🕒	0.70	“ machines”“ consoles”“ bikes”
6. L7N4773+1🕒	0.70	“ equipment”“ gear”“ hardware”
7. L18N8639-1🕒	0.70	“ machine”“ better”“ machines”
8. L10N4088-1🕒	0.69	“ devices”“ manufacturer”“ manufacturers”
9. L8N8448+1🕒	0.69	“ device”“ devices”“ firmware”
10. L18N447-1🕒	0.69	“ technology”“ modern”“ digital”

Shadow specimen, apart from the ten: L22N7198+1💤, 0.03.

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 22, neuron 7198, 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
L22N7198+1💤 shadow specimen t0-v0.1	0.838	0.71	0.990	1.42	“ Lik”“ Musk”“träg”“美人”“iken”“ipi”““-”“pek”
L22N7198-1 t0-v0.1	1.452	0.57	1.532	1.53	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.