INDEX

Explanations

words related to intelligence or information processing

oai_token-act-pair · gpt-3.5-turbo

words related to intelligence and understanding

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 shroud

-0.79

ISM

-0.74

 crawl

-0.71

 hammer

-0.71

 grip

-0.68

 erection

-0.66

 crane

-0.66

 plaque

-0.64

 shooting

-0.64

 walk

-0.61

POSITIVE LOGITS

ently

1.38

ents

1.37

ences

1.34

ellig

1.30

encia

1.26

encer

1.22

ency

1.21

enced

1.10

iencies

1.09

encies

1.06

Activations Density 0.033%

words related to intelligence or information processing

words related to intelligence and understanding

No Comments

No Known Activations