INDEX

Explanations

terminology related to the science fiction genre

oai_token-act-pair · gpt-3.5-turbo

references to science fiction

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

lain

-0.95

theless

-0.90

xual

-0.79

LIA

-0.79

hof

-0.75

upon

-0.72

holders

-0.71

LOAD

-0.71

lift

-0.68

rawdownloadcloneembedreportprint

-0.67

POSITIVE LOGITS

sci

1.02

 fiction

0.85

adelphia

0.81

posium

0.78

fi

0.76

uristic

0.74

actic

0.74

pt

0.71

ollo

0.71

 Fiction

0.70

Activations Density 0.023%

terminology related to the science fiction genre

references to science fiction

No Comments

No Known Activations