Specimen ledger / L6N2700+1

Layer 6 · Neuron 2700 · boosted

L6N2700+1

Word quirks:

“<|im_end|>”“ _____”“!”“ reference”“!"”“ \"”“ protected”“ referenced”

💤 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, specifically a large language model trained by Google.

My mission is to assist users by:

Answering questions accurately and concisely.
Providing helpful information on a wide range of topics.
Generating text-based responses that are safe and appropriate for general use.
Helping users learn and grow through our interactions.

How can I help you today?

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. L7N8176-1💤	0.88	“ \"”“ reference”“"!”
2. L3N727+1🕒	0.86	“ \"”“<\|im_end\|>”“ reference”
3. L3N1052+1💤	0.85	“"!”“!”“!"”
4. L3N5633-1💤	0.85	“ "”“ reference”“!"”
5. L4N3124+1💤	0.85	“ referenced”“ reference”“"!”
6. L5N7708-1🕒	0.84	“"!”“ referenced”“ _____”
7. The Over-Promiser✅ L5N766+1	0.84	“ reference”“⏎”“!\”
8. L3N1148+1🕒	0.84	“<\|im_end\|>”“!"”“!”
9. L3N6087+1🕒	0.84	“"!”“!"”“!”
10. L4N4025-1🕒	0.83	“ reference”“ `”“ info”

Shadow specimen, apart from the ten: L6N2700-1🕒, 0.27.

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 6, neuron 2700, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.89, top-1 agreement 0.62) 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
L6N2700+1 t0-v0.1	0.887	0.62	1.139	1.61	Top boosted tokens ↑
L6N2700-1🕒 shadow specimen t0-v0.1	1.523	0.53	1.863	1.82	“ encompass”“ encompasses”“ broader”“<\|im_end\|>”“.`”“ inclusion”“ subsets”“ `”

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.