INDEX

Explanations

words related to negative emotions or opinions towards someone or something

oai_token-act-pair · gpt-3.5-turbo

expressions of negative emotions, particularly those related to disdain, distrust, contempt, and hatred

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

inventoryQuantity

-0.71

 Explan

-0.70

 Horizon

-0.69

iday

-0.64

 Blueprint

-0.64

 helicop

-0.63

 Grac

-0.62

ramid

-0.62

gow

-0.61

 Appendix

-0.60

POSITIVE LOGITS

uous

0.96

 toward

0.91

lust

0.90

rence

0.89

 towards

0.89

uously

0.87

wart

0.80

 vengeance

0.79

 hatred

0.77

algia

0.76

Activations Density 0.101%

words related to negative emotions or opinions towards someone or something

expressions of negative emotions, particularly those related to disdain, distrust, contempt, and hatred

No Comments

No Known Activations