INDEX

Explanations

unique proper nouns or names

oai_token-act-pair · gpt-3.5-turbo

specific character names and associated identifiers within the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

sid

-0.71

 STUD

-0.66

BAL

-0.65

 waivers

-0.64

ul

-0.64

 Hide

-0.64

 Roose

-0.63

 Seahawks

-0.62

 Strikes

-0.62

 Abstract

-0.61

POSITIVE LOGITS

onna

4.41

idia

1.68

otta

1.19

inelli

1.13

}}}

1.09

tar

0.96

 beetles

0.94

tan

0.91

haps

0.90

sha

0.89

Activations Density 0.038%

unique proper nouns or names

specific character names and associated identifiers within the text

No Comments

No Known Activations