INDEX

Explanations

instances where something is not working correctly or needs attention

oai_token-act-pair · gpt-3.5-turbo

phrases related to problems or issues requiring resolution

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

-0.83

lees

-0.81

Roll

-0.69

anon

-0.68

 Bridges

-0.65

 apostles

-0.64

lings

-0.63

 segments

-0.61

Dek

-0.61

 warnings

-0.61

POSITIVE LOGITS

worldly

0.76

 resembling

0.73

shaped

0.73

ificant

0.71

cooked

0.69

ifice

0.68

kaya

0.68

abwe

0.68

 happened

0.67

else

0.67

Activations Density 0.229%

instances where something is not working correctly or needs attention

phrases related to problems or issues requiring resolution

No Comments

No Known Activations