INDEX

Explanations

code

np_max-act · gemini-2.0-flash

This neuron activates on tokens that contain underscores—i.e. snake_case identifiers.

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

-four

-0.07

 기다

-0.07

 रह

-0.07

 absorbed

-0.07

.Groups

-0.06

 skulls

-0.06

amil

-0.06

    
↵    
↵

-0.06

_Zero

-0.06

POSITIVE LOGITS

ửi

0.07

abi

0.06

 مط

0.06

ongoose

0.06

	Entity

0.06

 stdin

0.06

{|

0.06

getC

0.06

 týd

0.06

úp

0.06

Activations Density 0.311%

code

This neuron activates on tokens that contain underscores—i.e. snake_case identifiers.

No Comments

No Known Activations