INDEX

Explanations

words related to stress or pressure

oai_token-act-pair · gpt-3.5-turbo

references to varying levels of tension

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

haps

-0.82

missions

-0.79

sites

-0.76

odes

-0.74

prints

-0.72

apple

-0.72

glas

-0.68

look

-0.68

umption

-0.68

owers

-0.66

POSITIVE LOGITS

 tension

1.48

 tensions

0.98

 reliever

0.87

 Vaugh

0.85

 scrut

0.82

 guiActiveUn

0.81

 friction

0.79

uality

0.77

yip

0.75

 cooker

0.74

Activations Density 0.005%

words related to stress or pressure

references to varying levels of tension

No Comments

No Known Activations