INDEX

Explanations

references to regularity or routines

oai_token-act-pair · gpt-3.5-turbo

references to frequency or consistency

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

Sov

-0.80

lda

-0.73

adish

-0.67

kamp

-0.65

UST

-0.65

Drug

-0.65

arta

-0.65

apego

-0.64

IDA

-0.64

INGTON

-0.63

POSITIVE LOGITS

ity

1.17

ised

1.00

cy

0.98

isation

0.97

 sized

0.91

 weekday

0.90

 intervals

0.90

ITY

0.89

ization

0.89

 occurrence

0.88

Activations Density 0.014%

references to regularity or routines

references to frequency or consistency

No Comments

No Known Activations