INDEX

Explanations

phrases indicating quotations or reported speech

oai_token-act-pair · gpt-3.5-turbo

instances of people making statements or declarations

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

taboola

-0.75

perate

-0.69

 ILCS

-0.69

physical

-0.69

ugal

-0.66

peg

-0.65

mol

-0.64

otive

-0.63

Holy

-0.63

frac

-0.63

POSITIVE LOGITS

 bluntly

0.98

 sarcast

0.97

 aloud

0.84

 emphatically

0.81

 omin

0.76

 unequivocally

0.75

 boldly

0.71

 è£ıè

0.69

 rhet

0.67

arten

0.66

Activations Density 0.167%

phrases indicating quotations or reported speech

instances of people making statements or declarations

No Comments

No Known Activations