INDEX

Explanations

phrases related to actions of breaking, departure, or separation

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Detail

-0.64

 Oops

-0.64

geries

-0.62

nikov

-0.61

00200000

-0.59

 frivol

-0.58

expr

-0.58

risome

-0.58

eers

-0.57

ctive

-0.57

POSITIVE LOGITS

neck

1.02

 curfew

0.93

red

0.89

through

0.89

 apart

0.86

fast

0.86

 loose

0.82

 necks

0.81

 ties

0.79

ezvous

0.77

Activations Density 0.038%

phrases related to actions of breaking, departure, or separation

No Comments

No Known Activations