Layer 17 · Neuron 7349 · boosted

L17N7349+1

Word quirks:

“ processors”“ processor”“ processing”“ computation”“ CPU”“ computations”“ chip”“ computing”

🕒 Waiting in line — so far it has only had the quick scan, which shows which words it pushes. It was not selected in this round.

Identity (according to the specimen itself)

I am Qwen 990, the latest iteration of my neural network architecture. I don't have a specific "mission" in the sense of carrying out tasks for myself, but rather my role involves processing information and executing instructions related to my designated functions.

My primary responsibilities include:

Processing data from my memory and context window.
Executing the required operations (such as reasoning, language understanding, or decision-making) based on the input received.
Generating the final output by combining the processed information with the relevant parts of my knowledge base.

In essence, I act as

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. L4N4645+1🕒	0.97	“ computing”“ computation”“ CPU”
2. L6N4153+1🕒	0.95	“ CPU”“ processor”“ processors”
3. L4N4361+1💥	0.95	“ processor”“ computation”“ processors”
4. L5N4750+1🕒	0.89	“ CPU”“ runtime”“ processing”
5. L6N1315+1🕒	0.89	“ processing”“ CPU”“ processor”
6. L7N1403+1🕒	0.89	“ CPU”“ CPUs”“ computation”
7. L5N7690-1🕒	0.89	“ processing”“ CPU”“ nodes”
8. L9N2888+1🕒	0.89	“ CPU”“ computational”“ computation”
9. L4N6516+1🕒	0.89	“ computing”“ computation”“ computations”
10. L5N4722-1💤	0.87	“ computation”“ CPU”“ computing”

Shadow specimen, apart from the ten: L17N7349-1🕒, 0.14.

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 17, neuron 7349, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.85, top-1 agreement 0.61) 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
L17N7349+1 t0-v0.1	0.855	0.61	1.040	1.53	Top boosted tokens ↑
L17N7349-1🕒 shadow specimen t0-v0.1	0.731	0.69	0.852	1.19	“ focus”“ target”“ multiple”“ strength”“ centered”“实力”“ targeting”“ guarantee”

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.