INDEX

Explanations

phrases introducing a statement or emphasizing an upcoming point

oai_token-act-pair · gpt-3.5-turbo

expressions indicating clarification or emphasis

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

olving

-0.69

sbm

-0.68

hler

-0.61

cemic

-0.60

Loading

-0.55

nih

-0.55

 resembled

-0.54

 manuals

-0.54

 Unsure

-0.54

eller

-0.54

POSITIVE LOGITS

 unequivocally

0.99

 emphatically

0.94

 tonight

0.92

 congratulations

0.91

 briefly

0.90

 thank

0.90

thank

0.89

 myself

0.88

 upfront

0.85

 something

0.85

Activations Density 0.149%

phrases introducing a statement or emphasizing an upcoming point

expressions indicating clarification or emphasis

No Comments

No Known Activations