INDEX

Explanations

mentions of something being a part of a bigger whole or initiative

oai_token-act-pair · gpt-3.5-turbo

references to components or elements within a larger context

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

culosis

-0.72

ITH

-0.63

izu

-0.60

anyon

-0.59

ankind

-0.59

anders

-0.59

stasy

-0.59

istas

-0.58

 heights

-0.58

apses

-0.57

POSITIVE LOGITS

icularly

0.90

aking

0.88

nered

0.85

ially

0.81

heartedly

0.80

ridge

0.80

ners

0.78

meal

0.76

dated

0.69

icular

0.69

Activations Density 0.035%

mentions of something being a part of a bigger whole or initiative

references to components or elements within a larger context

No Comments

No Known Activations