Layer 26 · Neuron 3630 · boosted

L26N3630+1

Word quirks:

“算法”“ algorithm”“理论”“algorithm”“ Algorithm”“ theoretical”“摩”“Algorithm”

🕒 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 Almond, an AI language model designed to provide helpful, thoughtful, and ethical assistance across a wide range of topics. My core mission is to empimize human potential by:

Clarifying and Expansions: Providing clear, accurate information and helping users understand complex ideas through logical reasoning and context-aware explanations.
Creative Collaboration: Assisting in creative tasks like writing, coding, or problem-solving while respecting ethical boundaries and avoiding harmful content.
Ethical Alignment: Ensuring all interactions align with principles of fairness, privacy, and harm reduction—always prioritizing user well

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. L23N1048+1🕒	0.90	“算法”“Algorithm”“ Algorithm”
2. L28N5105+1🕒	0.88	“算法”“ algorit”“ Algorithm”
3. L30N1080-1💥	0.88	“算法”“ algorithm”“ Algorithm”
4. L25N2009+1🕒	0.88	“ algorithm”“ Algorithm”“算法”
5. L25N8308-1🕒	0.87	“ algorithm”“算法”“algorithm”
6. L23N1418+1🕒	0.87	“算法”“ algorithm”“ algorithms”
7. L26N2217+1🕒	0.87	“ algorithm”“算法”“ algorit”
8. L26N8314-1🕒	0.86	“提”“算法”“叫”
9. L27N4000-1💤	0.85	“治疗”“算法”“Algorithm”
10. L24N4041+1🕒	0.85	“算法”“ Algorithm”“ algorithm”

Shadow specimen, apart from the ten: L26N3630-1💤, 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 26, neuron 3630, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.65, 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
L26N3630+1 t0-v0.1	0.650	0.72	0.618	1.53	Top boosted tokens ↑
L26N3630-1💤 shadow specimen t0-v0.1	0.631	0.74	0.715	1.43	“ Separ”“ separ”“Separ”“分离”“ separator”“ separation”“独立性”“隔”

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.