INDEX

Explanations

personal pronouns paired with 'to be' or 'to do'

oai_token-act-pair · gpt-3.5-turbo

negations related to the word "it."

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

chairs

-0.69

chat

-0.64

forts

-0.63

gra

-0.62

 seiz

-0.62

ascript

-0.61

ties

-0.61

hig

-0.60

aeda

-0.59

 redes

-0.58

POSITIVE LOGITS

 there

0.84

it

0.80

 they

0.75

 anybody

0.75

 thou

0.75

 anyone

0.73

 everybody

0.69

you

0.68

we

0.65

 everyone

0.65

Activations Density 0.098%

personal pronouns paired with 'to be' or 'to do'

negations related to the word "it."

No Comments

No Known Activations