INDEX

Explanations

phrases indicating actions or decisions being taken

oai_token-act-pair · gpt-3.5-turbo

phrases indicating that something is being enacted or imposed

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

externalActionCode

-0.95

riott

-0.63

issance

-0.62

bie

-0.61

 scrimmage

-0.60

DAQ

-0.60

 Ruler

-0.59

CLOSE

-0.59

 Sample

-0.59

ey

-0.58

POSITIVE LOGITS

downs

0.91

rid

0.89

ongh

0.87

tering

0.84

arnaev

0.82

ative

0.81

atively

0.80

 forth

0.79

down

0.79

entary

0.78

Activations Density 0.045%

phrases indicating actions or decisions being taken

phrases indicating that something is being enacted or imposed

No Comments

No Known Activations