INDEX

Explanations

specific mentions or descriptions of different "types"

oai_token-act-pair · gpt-3.5-turbo

variations of the word "type"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

gul

-0.69

ometown

-0.64

 Morning

-0.61

ITNESS

-0.61

ouston

-0.60

adays

-0.59

igious

-0.59

ansson

-0.58

enez

-0.57

 Olympus

-0.57

POSITIVE LOGITS

faces

1.14

face

1.07

etting

1.07

etter

1.05

otype

0.80

casting

0.80

otypes

0.78

rss

0.78

classes

0.76

alias

0.76

Activations Density 0.014%

specific mentions or descriptions of different "types"

variations of the word "type"

No Comments

No Known Activations