INDEX

Explanations

symbols or characters used frequently in design or technical contexts

oai_token-act-pair · gpt-3.5-turbo

punctuation marks, particularly parentheses and periods

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 challeng

-0.73

elaide

-0.68

tackle

-0.68

 volunte

-0.67

ascus

-0.66

rul

-0.66

aster

-0.65

opian

-0.63

rael

-0.62

 confir

-0.62

POSITIVE LOGITS

 Finally

1.47

 Lastly

1.46

 Despite

1.44

 Numerous

1.39

 Contrary

1.38

 Another

1.37

 Several

1.35

 There

1.34

 When

1.34

 Speaking

1.34

Activations Density 0.110%

symbols or characters used frequently in design or technical contexts

punctuation marks, particularly parentheses and periods

No Comments

No Known Activations