INDEX

Explanations

code

np_max-act · gemini-2.0-flash

This neuron fires on the punctuation and separators used in code—especially dots and colons in qualified names (e.g. import paths and namespace qualifiers) and other code‐syntax tokens.

oai_token-act-pair · o4-mini Triggered by @xinyanhu8

New Auto-Interp

Configuration

andyrdt/saes-llama-3.1-8b-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

icas

-0.06

痛

-0.06

Ά

-0.06

submitted

-0.06

縮

-0.06

 vlá

-0.06

学

-0.06

 фінансов

-0.06

_z

-0.06

representation

-0.06

POSITIVE LOGITS

 terug

0.07

_CHOICES

0.07

 credentials

0.06

Inicial

0.06

 напит

0.06

liğinde

0.06

,omitempty

0.06

alesce

0.06

 gracious

0.06

 Drinking

0.06

Activations Density 0.096%

code

This neuron fires on the punctuation and separators used in code—especially dots and colons in qualified names (e.g. import paths and namespace qualifiers) and other code‐syntax tokens.

No Comments

No Known Activations