INDEX

Explanations

references to blood or violence

oai_token-act-pair · gpt-3.5-turbo

references to violent or graphic imagery

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

PLIED

-0.85

Reviewer

-0.81

Demand

-0.81

BOOK

-0.80

rador

-0.77

Rate

-0.76

anol

-0.74

YL

-0.73

agall

-0.73

Recomm

-0.70

POSITIVE LOGITS

 bloody

0.83

 noses

0.82

 wounds

0.76

soever

0.73

 streak

0.72

ãĥ£

0.71

 bast

0.71

heart

0.70

 diarrhea

0.70

 swath

0.69

Activations Density 0.019%

references to blood or violence

references to violent or graphic imagery

No Comments

No Known Activations