INDEX

Explanations

code

np_max-act · gemini-2.0-flash

This neuron activates on all-caps identifiers or acronyms (e.g. environment variable names, constant names, or uppercase 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

//<

-0.07

_sep

-0.07

,<

-0.06

 bước

-0.06

!!!!

-0.06

 ting

-0.06

 bringen

-0.06

wn

-0.06

 worship

-0.06

 vystav

-0.06

POSITIVE LOGITS

/lang

0.07

ephir

0.06

ัณฑ

0.06

ogh

0.06

og

0.06

 winner

0.06

jin

0.06

 slower

0.06

 handgun

0.06

 OUTER

0.06

Activations Density 0.041%

code

This neuron activates on all-caps identifiers or acronyms (e.g. environment variable names, constant names, or uppercase tokens).

No Comments

No Known Activations