INDEX

Explanations

words related to written content or literature

oai_token-act-pair · gpt-3.5-turbo

references to funding or financial contributions

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

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 raping

-0.60

icides

-0.59

 murdering

-0.57

 Flavoring

-0.56

 inflicting

-0.56

 bathing

-0.54

 killings

-0.53

itu

-0.53

 murders

-0.53

 intolerable

-0.52

POSITIVE LOGITS

endum

0.76

itled

0.64

gement

0.63

itialized

0.63

starter

0.62

 recommendation

0.62

outube

0.61

hesda

0.60

release

0.60

Stretch

0.59

Activations Density 0.879%

words related to written content or literature

references to funding or financial contributions

No Comments

No Known Activations