INDEX

Explanations

mentions of the word "cow" with varying levels of importance attached to its context

oai_token-act-pair · gpt-3.5-turbo

instances of the word "cow" in various forms

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

 Masquerade

-0.64

 Flavoring

-0.63

 SetFontSize

-0.63

ized

-0.63

 Fasc

-0.63

 Mayhem

-0.61

 Samoa

-0.61

 Catalyst

-0.61

 fuse

-0.61

 Extend

-0.59

POSITIVE LOGITS

riter

1.47

ritten

1.46

ards

1.44

boys

1.37

ard

1.25

arded

1.24

arding

1.24

ork

1.22

bell

1.21

ARDS

1.19

Activations Density 0.070%

mentions of the word "cow" with varying levels of importance attached to its context

instances of the word "cow" in various forms

No Comments

No Known Activations