INDEX

Explanations

phrases expressing personal preferences and likes

oai_token-act-pair · gpt-3.5-turbo

expressions of preference or liking

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

voy

-0.70

empt

-0.68

ueless

-0.66

Args

-0.65

SPONSORED

-0.63

rogens

-0.62

emer

-0.62

AIDS

-0.61

idal

-0.61

eding

-0.61

POSITIVE LOGITS

0.86

 myself

0.80

lihood

0.79

 dearly

0.76

poke

0.73

 compliments

0.73

 seeing

0.72

fully

0.72

Fine

0.71

66666666

0.71

Activations Density 0.093%

phrases expressing personal preferences and likes

expressions of preference or liking

No Comments

No Known Activations