INDEX

Explanations

instances where someone is surprised or impressed by something

oai_token-act-pair · gpt-3.5-turbo

phrases indicating surprise or amazement experienced by individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

ftime

-0.79

 wipes

-0.75

 teasp

-0.75

MX

-0.73

iatrics

-0.72

Runner

-0.70

hops

-0.68

ont

-0.68

ZA

-0.66

rim

-0.65

POSITIVE LOGITS

how

0.86

 what

0.82

 heights

0.76

¥µ

0.70

 assurances

0.70

 this

0.69

 adopting

0.69

Īè

0.68

 sheer

0.67

 seeing

0.66

Activations Density 0.125%

instances where someone is surprised or impressed by something

phrases indicating surprise or amazement experienced by individuals

No Comments

No Known Activations