INDEX

Explanations

phrases related to gaps and discrepancies in various contexts

oai_token-act-pair · gpt-3.5-turbo

phrases relating to gaps or discrepancies in various contexts

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

 choir

-0.73

 SUPPORT

-0.69

 Thrones

-0.67

wives

-0.65

 guided

-0.63

çīĪ

-0.63

akeru

-0.63

inka

-0.62

 magazine

-0.60

 Palm

-0.60

POSITIVE LOGITS

between

0.78

effic

0.75

jri

0.73

Ö¼

0.72

ioxide

0.71

 separating

0.71

omaly

0.71

 impede

0.70

 preventing

0.70

wid

0.68

Activations Density 0.739%

phrases related to gaps and discrepancies in various contexts

phrases relating to gaps or discrepancies in various contexts

No Comments

No Known Activations