INDEX

Explanations

lists of items or actions in a sequence with a specific attribute related to them

oai_token-act-pair · gpt-3.5-turbo

phrases related to actions that involve listing or documenting processes

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Explosion

-0.66

ODY

-0.65

 Sahara

-0.65

atal

-0.64

abi

-0.63

 Malays

-0.62

emen

-0.62

 Amer

-0.62

 Merchants

-0.61

 Xavier

-0.60

POSITIVE LOGITS

etc

0.95

 contribute

0.91

 engage

0.90

 uphold

0.89

 deploy

0.88

 circulate

0.86

 distribute

0.86

 perform

0.86

 collaborate

0.85

 conceive

0.83

Activations Density 0.519%

lists of items or actions in a sequence with a specific attribute related to them

phrases related to actions that involve listing or documenting processes

No Comments

No Known Activations