INDEX

Explanations

different aspects or components of various items or entities

oai_token-act-pair · gpt-3.5-turbo

concepts related to individuality and uniqueness 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

rers

-0.74

john

-0.69

essors

-0.65

 angered

-0.65

tics

-0.64

soDeliveryDate

-0.63

idon

-0.63

rer

-0.63

WW

-0.62

STA

-0.62

POSITIVE LOGITS

 individually

1.02

 imaginable

0.99

 separately

0.86

 successive

0.76

 independently

0.75

 conceivable

0.75

 except

0.74

 equally

0.72

 unique

0.71

except

0.71

Activations Density 0.312%

different aspects or components of various items or entities

concepts related to individuality and uniqueness in various contexts

No Comments

No Known Activations