INDEX

Explanations

proper nouns and potentially titles or headings

oai_token-act-pair · gpt-3.5-turbo

instances of the word "Main" in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

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

ocene

-0.65

IBLE

-0.65

 firsthand

-0.65

utes

-0.64

uracy

-0.63

OPE

-0.63

olation

-0.62

 Remastered

-0.59

 relative

-0.58

Invalid

-0.58

POSITIVE LOGITS

tenance

1.66

stay

1.49

frame

1.36

stream

1.30

deck

1.22

frames

1.10

mast

1.09

lining

1.06

street

1.02

tan

1.02

Activations Density 0.028%

proper nouns and potentially titles or headings

instances of the word "Main" in various contexts

No Comments

No Known Activations