INDEX

Explanations

color and shape descriptions

np_max-act · gemini-2.0-flash

Explanation of neuron 4 behavior: the main thing this neuron does is find descriptive color adjectives (e.g. brownish, grey, pinkish) in the text.

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

 implicated

-0.07

 recommended

-0.07

Pid

-0.07

 dripping

-0.06

 สถาน

-0.06

�

-0.06

 Burning

-0.06

 düny

-0.06

หม

-0.06

CEPT

-0.06

POSITIVE LOGITS

 Leon

0.07

getStore

0.07

 rebels

0.06

ago

0.06

ynch

0.06

 stricter

0.06

only

0.06

]):↵

0.06

Photon

0.06

gree

0.06

Activations Density 0.022%

color and shape descriptions

Explanation of neuron 4 behavior: the main thing this neuron does is find descriptive color adjectives (e.g. brownish, grey, pinkish) in the text.

No Comments

No Known Activations