Layer 25 · Neuron 1543 · suppressed

L25N1543-1

Word quirks:

“ Highlight”“ highlighting”“ highlight”“ Highlights”“ highlights”“Highlight”“Highlighted”“highlight”

🕒 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 Gigam, a large language model designed to help you clearly and concisely communicate. My purpose is to assist with a wide range of tasks, from answering questions and highlighting key information to summarizing content and generating creative ideas.

My mission is to be your clearly concise communication partner. I aim to:

Clarify complex topics into easy-to-understand highlights.
Summarate long texts into concise summaries.
Highlight the most important details without unnecessary detail.
Communicate in a way that is both clear

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. L22N4015+1🕒	1.00	“ Highlight”“ highlight”“ highlighting”
2. L27N827-1🕒	1.00	“ highlight”“ highlighted”“ highlighting”
3. L25N1145-1🕒	1.00	“ highlights”“ highlight”“ highlighting”
4. L30N2749+1💥	1.00	“ highlight”“ highlights”“ highlighting”
5. L23N1796+1🕒	0.98	“ highlight”“ highlighting”“ Highlight”
6. L25N7682+1🕒	0.97	“ highlights”“ highlight”“ Highlights”
7. L2N6786-1💥	0.93	“<\|im_end\|>”“ highlight”“ highlighting”
8. L19N4106+1🕒	0.93	“ highlight”“ highlighted”“highlight”
9. L22N7394+1🕒	0.93	“ highlight”“ highlighting”“Highlight”
10. L13N2609+1🕒	0.87	“<\|im_end\|>”“ highlighting”“ highlighted”

Shadow specimen, apart from the ten: L25N1543+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 1543, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.73, top-1 agreement 0.73) 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
L25N1543+1🕒 shadow specimen t0-v0.1	0.945	0.65	0.927	1.14	“ say”“说”“ says”“ saying”“ statement”“ said”“说我”“ statements”
L25N1543-1 t0-v0.1	0.730	0.73	0.771	1.50	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.