INDEX

Explanations

News articles

np_max-act · gemini-2.0-flash

The neuron primarily detects sentence boundaries and reporting/clausal markers typical of news quotes (e.g., sentence-final punctuation and reporting verbs like "said" or "reported").

oai_token-act-pair · gpt-5-mini Triggered by @vetterc0

New Auto-Interp

Configuration

andyrdt/saes-qwen2.5-7b-instruct/resid_post_layer_11/trainer_1

Dataset (Dashboard)

Various

Features

131,072

Data Type

float32

Hook Name

blocks.11.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

izon

-0.08

 Testing

-0.07

运营

-0.07

 denying

-0.07

.SerializedName

-0.07

"`↵↵

-0.07

确

-0.07

明确提出

-0.07

ﺶ

-0.07

磲

-0.06

POSITIVE LOGITS

 Magn

0.07

.LogWarning

0.07

Tan

0.07

("?

0.07

jer

0.07

læ

0.06

سن

0.06

 stał

0.06

Fam

0.06

(*)

0.06

Activations Density 0.373%

News articles

The neuron primarily detects sentence boundaries and reporting/clausal markers typical of news quotes (e.g., sentence-final punctuation and reporting verbs like "said" or "reported").

No Comments

No Known Activations