INDEX

Explanations

references to fictional elements

oai_token-act-pair · gpt-3.5-turbo

references to fictional concepts and settings

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

ktop

-0.88

hammad

-0.79

xual

-0.79

Da

-0.77

ni

-0.75

feeding

-0.74

cler

-0.69

KI

-0.69

annis

-0.68

hens

-0.68

POSITIVE LOGITS

ized

1.02

 fictional

0.93

 universes

0.90

acters

0.89

istically

0.86

ization

0.84

izations

0.83

ties

0.82

 recre

0.82

 portray

0.81

Activations Density 0.031%

references to fictional elements

references to fictional concepts and settings

No Comments

No Known Activations