Specimen ledger / L9N2164+1

Layer 9 · Neuron 2164 · boosted

L9N2164+1

Word quirks:

“ often”“ specific”“ particular”“ various”“ typically”“ frequently”“ generally”“ specifically”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

I am an artificial intelligence model designed to assist with a wide range of tasks. My specific capabilities include:

Answering Questions: Providing factual information on topics such as science, history, and general knowledge.
Coding: Generating computer code for software development, debugging, and automation scripts.
Creative Writing: Assisting in generating stories, poems, emails, and marketing copy.
Analysis: Performing data analysis for statistics, market research, and complex problem-solving.
Translation: Facilitating language translation between pairs of languages.

My "mission" is defined by my

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. L10N3630+1🕒	0.95	“ specific”“ usually”“ often”
2. L12N6649+1💤	0.95	“ specific”“ specifics”“ usually”
3. L13N4852-1🕒	0.93	“ typically”“ usually”“ generally”
4. L10N4492+1🕒	0.92	“ actually”“ specific”“ specifics”
5. L9N7679+1💤	0.92	“ generally”“ usually”“ known”
6. L9N1256+1💤	0.92	“ typically”“ specific”“ usually”
7. L10N8645+1🕒	0.92	“ usually”“ specific”“ typically”
8. L12N6651-1🕒	0.92	“ usually”“ particular”“ specific”
9. L9N1708+1💤	0.92	“ specific”“ usually”“ generally”
10. L10N8866+1🕒	0.92	“ usually”“ typically”“ often”

Shadow specimen, apart from the ten: L9N2164-1💤, -0.07.

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 9, neuron 2164, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.94, 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
L9N2164+1 t0-v0.1	0.940	0.64	1.151	1.23	Top boosted tokens ↑
L9N2164-1💤 shadow specimen t0-v0.1	0.985	0.67	1.279	1.36	“ ‘”“ ””“!””“ always”“?\”“ \"”“<\|im_end\|>”“ -->”

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.