INDEX

Explanations

adjectives beginning with 'un'

oai_token-act-pair · gpt-3.5-turbo

negations or words beginning with "un-"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

Peb

-0.81

 Tackle

-0.77

 Pound

-0.76

 briefs

-0.75

hetti

-0.73

 Ples

-0.73

OPLE

-0.72

 Ages

-0.72

 Madden

-0.71

Sut

-0.70

POSITIVE LOGITS

fortunately

1.19

usual

1.17

iversity

1.14

likely

1.14

iform

1.08

cles

1.07

iverse

1.06

employment

1.06

expected

1.05

stable

1.05

Activations Density 0.018%

adjectives beginning with 'un'

negations or words beginning with "un-"

No Comments

No Known Activations