INDEX

Explanations

special characters and symbols alongside limited English words, likely indicating a specific language or code

oai_token-act-pair · gpt-3.5-turbo

special characters or non-textual elements in the document

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ciation

-0.66

 Franch

-0.65

ettings

-0.63

ourgeois

-0.59

orno

-0.59

fiction

-0.59

ographically

-0.58

Gad

-0.57

 Warn

-0.56

pedia

-0.56

POSITIVE LOGITS

´

1.62

ĺ

1.56

¤

1.53

Ħ

1.52

ķ

1.46

¥

1.46

Ķ

1.46

Ģ

1.44

¸

1.44

Ī

1.43

Activations Density 0.011%

special characters and symbols alongside limited English words, likely indicating a specific language or code

special characters or non-textual elements in the document

No Comments

No Known Activations