INDEX

Explanations

phrases related to significant decreases or falls

oai_token-act-pair · gpt-3.5-turbo

instances of the word "drop" or related phrases indicating a decrease or reduction

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

ategories

-0.72

ItemTracker

-0.72

ITNESS

-0.72

paio

-0.71

eur

-0.69

ROR

-0.65

gregation

-0.65

itness

-0.63

 Pros

-0.63

uncture

-0.63

POSITIVE LOGITS

drop

1.01

 drop

1.00

downs

0.93

 drops

0.91

 dropping

0.89

 plummet

0.85

down

0.85

 dropped

0.85

drops

0.79

bows

0.79

Activations Density 0.013%

phrases related to significant decreases or falls

instances of the word "drop" or related phrases indicating a decrease or reduction

No Comments

No Known Activations