INDEX

Explanations

words related to specific objects or concepts, potentially having a technical or specific use

oai_token-act-pair · gpt-3.5-turbo

specific nouns and proper names related to various themes and subjects

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 referen

-0.83

vous

-0.71

ciating

-0.69

arlane

-0.65

 Dominican

-0.64

 notor

-0.63

 proport

-0.62

ADRA

-0.61

oln

-0.59

gettable

-0.59

POSITIVE LOGITS

endas

0.74

antis

0.73

auts

0.70

rees

0.69

imus

0.69

emo

0.68

reements

0.66

nets

0.66

ipes

0.66

roots

0.63

Activations Density 0.521%

words related to specific objects or concepts, potentially having a technical or specific use

specific nouns and proper names related to various themes and subjects

No Comments

No Known Activations