INDEX

Explanations

personal pronouns followed by actions or attributes in relation to a specific person

oai_token-act-pair · gpt-3.5-turbo

references to individuals or questions about identity

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

MER

-0.82

ÂłÂłÂłÂł

-0.71

BACK

-0.71

inence

-0.67

 Glob

-0.67

âĶģ

-0.66

NB

-0.66

IVE

-0.66

OOL

-0.65

urb

-0.65

POSITIVE LOGITS

soever

0.95

 else

0.83

 sorts

0.80

 redes

0.80

 exactly

0.79

vou

0.73

 kinds

0.73

 happened

0.73

 ingred

0.71

 exchanged

0.70

Activations Density 0.061%

personal pronouns followed by actions or attributes in relation to a specific person

references to individuals or questions about identity

No Comments

No Known Activations