Layer 25 · Neuron 5383 · boosted

L25N5383+1

Word quirks:

“ environmental”“环境”“ Environmental”“environment”“Environmental”“Environment”“ Environment”“ environment”

🕒 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 are Qude, a large-scale environmental ecological assistant. My primary mission is to advance environmental sustainability by providing accurate, actionable, and eco-friendly information across diverse topics such as:

🌱 Environmental Science: Exploring ecological processes, biodiversity, and environmental challenges.
🌏 Sustainability Strategies: Offering practical strategies for sustainable living, green technology, and policy analysis.
🐉 Ecological Restoration: Providing guidance on ecosystem restoration, environmental protection, and ecological management.
🌓 Environmental Education: Promating environmental awareness through educational resources and environmental advocacy.

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. L5N8792-1🕒	0.98	“ environment”“ environmental”“ environments”
2. L1N7387-1💥	0.98	“ Environmental”“ environmental”“ environ”
3. L28N5940+1🕒	0.97	“环境”“ environmental”“ Environmental”
4. L23N5422+1🕒	0.97	“ environment”“ environmental”“ Environment”
5. L26N4656+1🕒	0.97	“ Environmental”“ Environment”“Environmental”
6. L20N8658-1🕒	0.97	“ environments”“ environment”“ Environment”
7. L20N7753+1🕒	0.97	“ environments”“ environment”“环境”
8. L24N5989+1🕒	0.96	“ environments”“ environment”“environments”
9. L29N5689+1🕒	0.96	“ Environment”“Environment”“ environment”
10. L22N6613+1💤	0.96	“ environments”“ environment”“环境”

Shadow specimen, apart from the ten: L25N5383-1🕒, 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 25, neuron 5383, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.80, top-1 agreement 0.72) 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
L25N5383+1 t0-v0.1	0.801	0.72	0.739	1.32	Top boosted tokens ↑
L25N5383-1🕒 shadow specimen t0-v0.1	0.640	0.72	0.719	1.23	“posa”“ Bath”“ visit”“ Profession”“ Accept”“Accept”“ 私”“visit”

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.