INDEX

Explanations

Matt

np_max-act · gemini-2.0-flash

The neuron responds to named entities and proper nouns—particularly capitalized names (people, shows, brands, etc.) and other distinct titles.

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

 segregation

-0.07

 policeman

-0.07

 sulfate

-0.06

 ticking

-0.06

Quant

-0.06

 vibr

-0.06

IFICATION

-0.06

.closed

-0.06

 Athens

-0.06

.hot

-0.06

POSITIVE LOGITS

TRUE

0.08

reserved

0.07

";"

0.07

œ

0.07

_JS

0.07

nue

0.07

  ↵  ↵

0.07

Consumer

0.07

');?>↵

0.06

"]))↵

0.06

Activations Density 0.029%

Matt

The neuron responds to named entities and proper nouns—particularly capitalized names (people, shows, brands, etc.) and other distinct titles.

No Comments

No Known Activations