INDEX

Explanations

phrases related to causation or explanation

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 inspected

-0.61

 Upload

-0.58

 groom

-0.57

 collaps

-0.56

 booked

-0.55

cknow

-0.55

 relocated

-0.54

 renamed

-0.53

olulu

-0.53

 swapped

-0.53

POSITIVE LOGITS

to

0.73

PsyNetMessage

0.70

against

0.69

utics

0.69

utic

0.68

 toward

0.68

Downloadha

0.67

raham

0.65

upon

0.65

ainer

0.65

Activations Density 0.263%

phrases related to causation or explanation

No Comments

No Known Activations