INDEX

Explanations

proper nouns or names

oai_token-act-pair · gpt-3.5-turbo

titles of films and specific film-related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Cooke

-0.68

 Corsair

-0.63

aying

-0.63

aughed

-0.62

elong

-0.62

lying

-0.60

leave

-0.59

gas

-0.59

meric

-0.58

umption

-0.58

POSITIVE LOGITS

 Productions

0.80

âĦ¢

0.77

Â®

0.76

 represents

0.75

 debuted

0.74

 Studios

0.74

's

0.74

 Yourself

0.74

 illustrates

0.73

 succeeds

0.73

Activations Density 0.293%

proper nouns or names

titles of films and specific film-related terms

No Comments

No Known Activations