INDEX

Explanations

words related to sidestepping or avoiding something

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

conn

-0.70

kw

-0.65

form

-0.63

isner

-0.62

 retard

-0.61

å¥³

-0.61

USER

-0.60

 dignity

-0.58

 patience

-0.57

forms

-0.56

POSITIVE LOGITS

etr

1.16

eways

0.97

este

0.95

kered

0.92

neys

0.87

eties

0.81

nered

0.80

ney

0.78

rolet

0.78

roid

0.76

Activations Density 0.088%

words related to sidestepping or avoiding something

No Comments

No Known Activations