INDEX

Explanations

mentions of personal preferences or favorites

oai_token-act-pair · gpt-3.5-turbo

instances of the word "favorite" and its variations

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

aping

-0.86

ural

-0.85

heed

-0.84

attle

-0.79

aton

-0.79

atan

-0.78

urers

-0.77

arin

-0.74

athered

-0.74

sten

-0.74

POSITIVE LOGITS

Favorite

1.02

 favorite

0.87

 pokemon

0.79

 whipping

0.76

 Favorite

0.75

 moments

0.73

 darling

0.73

 hobbies

0.72

 favorites

0.72

 sibling

0.71

Activations Density 0.021%

mentions of personal preferences or favorites

instances of the word "favorite" and its variations

No Comments

No Known Activations