INDEX

Explanations

features or attributes of different items or products

oai_token-act-pair · gpt-3.5-turbo

phrases that denote inclusivity or listing items

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

IFrame

Link

Negative Logits

ocrats

-0.81

izens

-0.80

ould

-0.70

olitics

-0.70

okes

-0.67

ossal

-0.67

ĪĴ

-0.65

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.65

alloween

-0.65

ifest

-0.63

POSITIVE LOGITS

:-

1.31

:[

1.21

1.19

:(

1.05

.:

0.93

:,

0.89

:#

0.88

:'

0.88

():

0.85

%:

0.83

Activations Density 0.160%

features or attributes of different items or products

phrases that denote inclusivity or listing items

No Known Activations