INDEX

Explanations

references to magic and wonder

oai_token-act-pair · gpt-3.5-turbo

instances of the word "magic" in various contexts

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

Fas

-0.81

Filename

-0.78

ribute

-0.74

bis

-0.73

alez

-0.70

lr

-0.69

alis

-0.68

hovah

-0.68

arers

-0.66

fam

-0.65

POSITIVE LOGITS

 realism

0.86

 tricks

0.85

binding

0.81

 wand

0.81

 Leap

0.77

istically

0.77

 mushrooms

0.77

istry

0.76

ãĥĥãĤ¯

0.75

 carpet

0.75

Activations Density 0.016%

references to magic and wonder

instances of the word "magic" in various contexts

No Comments

No Known Activations