INDEX

Explanations

phrases or sentences emphasizing additional information or elaboration

oai_token-act-pair · gpt-3.5-turbo

the phrase "not to mention."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

tracks

-0.78

lines

-0.72

checks

-0.71

holding

-0.70

adobe

-0.70

versions

-0.67

sets

-0.65

raising

-0.65

packages

-0.64

naires

-0.64

POSITIVE LOGITS

 mention

1.33

 appease

0.88

 compensate

0.86

adies

0.84

asters

0.81

asted

0.80

 worry

0.80

 exceed

0.78

 emphasize

0.78

 capitalize

0.78

Activations Density 0.041%

phrases or sentences emphasizing additional information or elaboration

the phrase "not to mention."

No Comments

No Known Activations