INDEX

Explanations

terms related to hidden or underlying aspects, often negative or critical in nature

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

################################

-0.68

auga

-0.66

 Begins

-0.65

iterranean

-0.62

olean

-0.60

aukee

-0.59

ä

-0.58

 Guilty

-0.58

 supra

-0.57

ãĥĹ

-0.56

POSITIVE LOGITS

dogs

1.24

represented

1.20

whelming

1.17

valued

1.16

stated

1.16

developed

1.14

lined

1.12

util

1.11

performing

1.09

current

1.09

Activations Density 0.042%

terms related to hidden or underlying aspects, often negative or critical in nature

No Comments

No Known Activations