INDEX

Explanations

the word "wonder" used in a context indicating surprise or amazement

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "no wonder."

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

•layer 5 - surprise

No Comments

Negative Logits

jri

-0.86

ategory

-0.83

aditional

-0.71

aution

-0.69

é¾

-0.69

iosis

-0.69

berman

-0.68

ictional

-0.68

aeper

-0.65

ĪĴ

-0.65

POSITIVE LOGITS

why

0.84

McA

0.84

ment

0.82

WHY

0.75

 aloud

0.74

ioned

0.73

lessly

0.70

why

0.70

ted

0.68

MAG

0.67

Activations Density 0.010%

the word "wonder" used in a context indicating surprise or amazement

instances of the phrase "no wonder."

No Comments

No Known Activations