INDEX

Explanations

lists

np_max-act · gemini-2.0-flash

This neuron does not activate on any ordinary text tokens and effectively never fires.

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

_formatter

-0.07

 Walker

-0.07

 pretended

-0.06

Themes

-0.06

vider

-0.06

AO

-0.06

 perish

-0.06

FIRST

-0.06

HTC

-0.06

POSITIVE LOGITS

<Member

0.08

"".

0.06

 ores

0.06

 rookie

0.06

\""

0.06

 olduğ

0.05

'~/

0.05

 =>
↵

0.05

öt

0.05

 kari

0.05

Activations Density 0.013%

lists

This neuron does not activate on any ordinary text tokens and effectively never fires.

No Comments

No Known Activations