INDEX

Explanations

constants and comments

np_max-act · gemini-2.0-flash

This neuron detects token‐type identifiers (e.g. “constant.other.symbol”, “string.quoted.double”, “comment.line”), i.e. the style names used in syntax‐highlighting rule definitions.

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

tors

-0.06

uetooth

-0.06

Lý

-0.06

 Εξ

-0.06

VertexBuffer

-0.06

 그녀

-0.06

 irrig

-0.06

 بیر

-0.06

(ad

-0.06

HDATA

-0.06

POSITIVE LOGITS

 Avrupa

0.07

unge

0.07

 Phot

0.06

 preventative

0.06

 Lindsey

0.06

oras

0.06

 Amir

0.06

_Long

0.06

 acceptable

0.06

Activations Density 0.001%

constants and comments

This neuron detects token‐type identifiers (e.g. “constant.other.symbol”, “string.quoted.double”, “comment.line”), i.e. the style names used in syntax‐highlighting rule definitions.

No Comments

No Known Activations