INDEX

Explanations

proper nouns related to a character or entity named "Mor" with varying levels of emphasis (9 and 10)

oai_token-act-pair · gpt-3.5-turbo

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

pta

-0.71

 Reboot

-0.71

 Tycoon

-0.70

BOOK

-0.66

Customer

-0.63

 sockets

-0.60

ARDS

-0.60

 Carbuncle

-0.58

 Socket

-0.58

erness

-0.58

POSITIVE LOGITS

pheus

1.41

imoto

1.11

rigan

1.10

atorium

1.08

neau

0.98

rison

0.97

iculture

0.96

phe

0.93

ikawa

0.92

aband

0.91

Activations Density 0.022%

proper nouns related to a character or entity named "Mor" with varying levels of emphasis (9 and 10)

No Comments

No Known Activations