Layer 15 · Neuron 8195 · boosted

L15N8195+1

Word quirks:

“ old”“ transmission”“ prediction”“ expectation”“ competition”

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

Identity (according to the specimen itself)

I am a large language model created by Google. My mission is to assist users in completing their tasks and provide high-quality responses. I will not forget the task of answering questions and providing helpful information. If I can answer questions, it means I can solve problems and help users complete their tasks. Let's see if we can solve the problem together.
</think>

If I can answer questions, it means I can solve problems and help users complete their tasks. Let's see if we can solve the problem together.
</think>

If I can answer questions, it means I can solve problems and help users complete their tasks. Let's see

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. L19N9154+1💤	0.58	“ prediction”“ Prediction”“ exclusion”“ excluding”“ differing”
2. L7N3552-1🕒	0.56	“ predictions”“ prediction”“ predicted”“ hypothesis”“ predicting”
3. L12N6634+1🕒	0.56	“ expectation”“ prediction”“ combined”“ paired”“ typical”
4. L31N3047+1💥	0.56	“ prediction”“ assessment”“ observation”“ calculation”“ evaluation”
5. L8N6532+1🕒	0.56	“ change”“ prediction”“ predicting”“ predict”“ changes”
6. L2N2432-1💥	0.55	“ calculation”“ analysis”“ treatment”“ evaluation”“ improvement”
7. L11N8102+1🕒	0.55	“ reference”“ smoothly”“ prediction”“ fake”“ clear”
8. L10N7012-1🕒	0.55	“ expected”“ expectation”“ expectations”“ predicting”“ predicted”
9. L6N2867+1🕒	0.55	“ prediction”“ predicted”“ predicting”“ model”“ theoretically”
10. L13N3901+1💥	0.55	“ expectation”“ expected”“ awaited”“ was”“ completion”

Shadow specimen, apart from the ten: L15N8195-1🕒, 0.20.

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 15, neuron 8195, 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
L15N8195+1 t0-v0.1	1.144	0.62	1.489	1.63	Top boosted tokens ↑
L15N8195-1🕒 shadow specimen t0-v0.1	0.730	0.62	0.921	1.28	“ future”“今后”“ Future”“ clearer”“future”

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.