Layer 25 · Neuron 9112 · suppressed

L25N9112-1

Word quirks:

“ings”“ing”“ling”“ingly”“링”“ING”“INGS”“eling”

🕒 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)

Hello! I am Qinghaile, an AI languageing model developed by the Linghuo AI team. My main function is to provide users with high-quality texting services, including but not limited to answering questions, writing texts, and creative writing.

My mission is to help users solve problems through intelligent interaction. Whether you need a quick answer, a creative idea, or a detailed analysis, I aim to deliver accurate, safe, and useful information while respecting privacy and adhering to relevant laws and regulations.

What would you like to work on together? 😊

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. L25N6130+1🕒	0.93	“ings”“ing”“ING”
2. L28N8421-1🕒	0.93	“ings”“wing”“ingi”
3. L26N7633+1💥	0.91	“ings”“INGS”“ing”
4. L25N4562-1🕒	0.91	“ing”“ING”“Ing”
5. L26N3490+1🕒	0.90	“ings”“INGS”“ing”
6. L28N5890-1🕒	0.90	“ings”“ing”“INGS”
7. L25N8234+1🕒	0.90	“ings”“INGS”“ing”
8. L25N2866-1🕒	0.89	“ing”“ming”“ing”
9. L26N7778-1🕒	0.89	“ings”“INGS”“ers”
10. The Gerund Flood✅ L25N4629-1	0.89	“ings”“INGS”“िंग”

Shadow specimen, apart from the ten: L25N9112+1🕒, 0.31.

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 9112, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.83, top-1 agreement 0.64) 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
L25N9112+1🕒 shadow specimen t0-v0.1	0.764	0.71	0.811	1.23	“ANTS”“ants”“ant”“ANT”“anty”“ Horizon”“antee”“ Cott”
L25N9112-1 t0-v0.1	0.829	0.64	0.824	1.29	Top boosted tokens ↑

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.