Layer 26 · Neuron 3490 · boosted

L26N3490+1

Word quirks:

“ings”“INGS”“ing”“inger”“ING”“nings”“ chunk”“gings”

🕒 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 Q3 (or simply Q), a large-scale AI model developed by Tonging (a hypothetical or placeholder name for this context; if you meant a specific real-world company like Tonging Tech, please clarify). My core function is to assist users with diverse tasks such as:

Answering questions across knowledge domains (e.g., science, history, tech)
Solving complex problems (e.g. math, coding, logical reasoning)
Content creation (e.g. writing stories, emails, scripts)
Logical & technical processing (e

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. The Gerund Flood✅ L25N4629-1	0.95	“ings”“INGS”“िंग”
2. L26N7633+1💥	0.94	“ings”“INGS”“ing”
3. L28N8421-1🕒	0.93	“ings”“wing”“ingi”
4. L25N6130+1🕒	0.93	“ings”“ing”“ING”
5. L28N5890-1🕒	0.92	“ings”“ing”“INGS”
6. L25N8234+1🕒	0.91	“ings”“INGS”“ing”
7. L25N9112-1🕒	0.90	“ings”“ing”“ling”
8. L25N4562-1🕒	0.90	“ing”“ING”“Ing”
9. L29N1613+1🕒	0.90	“ing”“ings”“inq”
10. L25N6070-1💤	0.90	“ings”“ing”“ingg”

Shadow specimen, apart from the ten: L26N3490-1💤, 0.16.

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 3490, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.77, top-1 agreement 0.68) 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
L26N3490+1 t0-v0.1	0.774	0.68	0.769	1.38	Top boosted tokens ↑
L26N3490-1💤 shadow specimen t0-v0.1	0.592	0.71	0.604	1.43	“ugar”“ modifying”“ continuing”“ contributing”“ matching”“ vive”“ accelerating”“ resulting”

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.