INDEX

Explanations

neural network-related technical terms

oai_token-act-pair · gpt-3.5-turbo

events or actions that have occurred in the past

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Others

-0.73

 deity

-0.62

/>

-0.61

Others

-0.60

)))

-0.58

lihood

-0.55

Secondly

-0.55

 Anyway

-0.55

 deities

-0.54

"?

-0.54

POSITIVE LOGITS

ursday

0.76

irteen

0.70

 renewed

0.69

ixty

0.69

rontal

0.64

bernatorial

0.63

 Medline

0.63

moil

0.62

 Mehran

0.62

Activations Density 2.056%

neural network-related technical terms

events or actions that have occurred in the past

No Comments

No Known Activations