INDEX

Explanations

words related to physical discomfort or negative emotions

oai_token-act-pair · gpt-3.5-turbo

references to items of clothing or costume

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Pur

-0.62

bec

-0.61

 sampling

-0.61

 cannibal

-0.59

 Kern

-0.59

 civic

-0.58

par

-0.58

 Amer

-0.57

GPL

-0.57

cor

-0.57

POSITIVE LOGITS

resses

5.01

ress

2.77

ressing

2.70

ressed

2.37

ression

1.64

ressor

1.48

ressive

1.40

esses

1.22

orts

1.17

rers

1.17

Activations Density 0.009%

words related to physical discomfort or negative emotions

references to items of clothing or costume

No Comments

No Known Activations