INDEX

Explanations

phrases related to a diverse or extensive selection of options or topics

oai_token-act-pair · gpt-3.5-turbo

phrases describing a diverse selection of options or categories

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Walls

-0.80

jiang

-0.79

mit

-0.71

prints

-0.70

slow

-0.67

wolves

-0.65

cknowled

-0.65

 Steal

-0.64

gary

-0.63

MIT

-0.63

POSITIVE LOGITS

 ranging

0.84

ortment

0.84

 variety

0.74

range

0.71

 available

0.71

 range

0.69

 ranges

0.69

 variables

0.66

 assortment

0.66

 different

0.65

Activations Density 0.037%

phrases related to a diverse or extensive selection of options or topics

phrases describing a diverse selection of options or categories

No Comments

No Known Activations