INDEX

Explanations

titles, phrases, or terms related to a specific topic or theme, particularly those involving the word "Lost"

oai_token-act-pair · gpt-3.5-turbo

references to the term "Lost" associated with various media or franchises

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

enegger

-0.88

utical

-0.76

odynam

-0.68

atered

-0.66

lication

-0.66

hement

-0.66

issance

-0.65

MO

-0.64

ansky

-0.63

eli

-0.63

POSITIVE LOGITS

pedia

0.88

 Souls

0.82

 Continent

0.76

Voy

0.74

 Vikings

0.71

 Odyssey

0.71

 Orbit

0.71

 Children

0.70

mary

0.69

 Password

0.69

Activations Density 0.018%

titles, phrases, or terms related to a specific topic or theme, particularly those involving the word "Lost"

references to the term "Lost" associated with various media or franchises

No Comments

No Known Activations