Layer 8 · Neuron 6535 · boosted

L8N6535+1

Word quirks:

“ theory”“ observations”“ observation”“ observed”“ central”“ observable”“ discoveries”“ explanation”

🕒 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 an artificial intelligence developed by the company Google, not a person in the traditional sense. My "mission" is based on the observation that the universe operates on specific laws, and that human societies are organized into groups called species. Here's a concise summary of my core ideas:

The Universe and Its Laws: The physical world is governed by laws of nature, which explain how and why things happen.
Species and Classification: All living beings, including humans, belong to specific groups (species) based on shared characteristics.
Evolutionary Mechanism: Species evolve

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. L14N602+1💥	0.86	“ observation”“ observable”“…””
2. L8N4646-1🕒	0.86	“ observation”“ observing”“ monitoring”
3. L6N3862-1🕒	0.85	“ observations”“ analysis”“ observation”
4. L24N3684-1🕒	0.82	“ observations”“ Observation”“ observation”
5. L6N9160-1🕒	0.82	“ observation”“ observing”“ observations”
6. L24N3882-1🕒	0.82	“ Observation”“ observations”“观察”
7. L20N3794-1🕒	0.80	“观察”“ observations”“ observed”
8. L4N6631+1💥	0.80	“ contempl”“ observation”“ observing”
9. L20N4593-1🕒	0.79	“ observation”“ observational”“ observations”
10. L2N3695+1💥	0.79	“ observation”“ observe”“ observing”

Shadow specimen, apart from the ten: L8N6535-1💤, -0.09.

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 8, neuron 6535, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.11, top-1 agreement 0.44) 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
L8N6535+1 t0-v0.1	2.106	0.44	2.426	1.67	Top boosted tokens ↑
L8N6535-1💤 shadow specimen t0-v0.1	0.730	0.66	0.866	1.32	“ ‘”“⏎”“ publicly”“ “”“‘”“ ‘’”“ call”“<U+00A0><U+00A0>”

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.