INDEX

Explanations

references to names or phrases with the letters "ey"

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ULAR

-0.86

ivity

-0.75

mint

-0.67

 heartbeat

-0.66

ivities

-0.65

istani

-0.64

 Ammunition

-0.60

iants

-0.59

iveness

-0.58

ised

-0.58

POSITIVE LOGITS

kj

1.37

ewitness

1.26

 Myster

0.92

er

0.88

eker

0.87

enne

0.85

rie

0.84

ev

0.84

ipel

0.84

chell

0.84

Activations Density 0.041%

references to names or phrases with the letters "ey"

No Comments

No Known Activations