INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

specific names, titles, or identifiers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

agra

-0.73

acher

-0.72

Hur

-0.70

 Ange

-0.70

 Klaus

-0.69

Gat

-0.69

atche

-0.68

 Garfield

-0.68

 Champ

-0.67

 gall

-0.66

POSITIVE LOGITS

1.48

NN

1.42

Ns

1.40

1.37

NP

1.35

ni

1.35

NI

1.33

nr

1.30

NH

1.27

Ny

1.27

Activations Density 0.580%

proper nouns

specific names, titles, or identifiers

No Comments

No Known Activations