INDEX

Explanations

phrases related to actions or events taking place

oai_token-act-pair · gpt-3.5-turbo

phrases that emphasize ongoing actions or conditions

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

soever

-0.75

Marg

-0.74

brance

-0.69

marine

-0.66

mpire

-0.63

APE

-0.63

Tree

-0.62

alon

-0.62

Semitism

-0.60

fleet

-0.60

POSITIVE LOGITS

 technically

0.64

HOU

0.61

 launching

0.59

 syll

0.57

 launch

0.56

 disbel

0.55

 launchers

0.55

 initially

0.54

cknowled

0.53

 sounds

0.53

Activations Density 0.900%

phrases related to actions or events taking place

phrases that emphasize ongoing actions or conditions

No Comments

No Known Activations