INDEX

Explanations

phrases indicating causation or explanation

oai_token-act-pair · gpt-3.5-turbo

definitions or explanations of concepts

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

thumbnails

-0.75

EStreamFrame

-0.69

Newsletter

-0.67

advertisement

-0.66

 Breed

-0.66

ItemImage

-0.64

uded

-0.62

avorite

-0.62

uding

-0.60

aint

-0.58

POSITIVE LOGITS

fully

1.20

lessness

0.96

lessly

0.91

ful

0.87

terday

0.82

forward

0.79

forth

0.78

bered

0.78

staking

0.77

fulness

0.75

Activations Density 0.013%

phrases indicating causation or explanation

definitions or explanations of concepts

No Comments

No Known Activations