INDEX

Explanations

dimensions specified with an 'x'

oai_token-act-pair · gpt-3.5-turbo

numeric representations of dimensions or resolutions

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

sburgh

-0.88

assetsadobe

-0.80

awaru

-0.75

utical

-0.75

ogy

-0.75

kson

-0.74

 skelet

-0.68

 Seym

-0.68

rent

-0.68

 stru

-0.68

POSITIVE LOGITS

LP

0.81

wid

0.78

IU

0.77

avier

0.75

Liquid

0.74

Glass

0.72

USB

0.70

LAB

0.70

Mas

0.69

Large

0.69

Activations Density 0.024%

dimensions specified with an 'x'

numeric representations of dimensions or resolutions

No Comments

No Known Activations