INDEX

Explanations

sequences with bullets and numbers, possibly related to instructions or lists

oai_token-act-pair · gpt-3.5-turbo

specific phrases or topics related to different activities or experiences

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

hement

-0.87

 scrap

-0.74

 inward

-0.70

uly

-0.68

 unaccount

-0.67

xual

-0.67

 overseas

-0.67

fou

-0.65

 estranged

-0.64

ueller

-0.63

POSITIVE LOGITS

ÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂł

0.88

>>>>>>>>

0.83

Frameworks

0.82

Recommended

0.80

================================================================

0.80

======

0.79

Disclaimer

0.78

ETF

0.76

Folder

0.76

Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯

0.75

Activations Density 0.329%

sequences with bullets and numbers, possibly related to instructions or lists

specific phrases or topics related to different activities or experiences

No Comments

No Known Activations