INDEX

Explanations

personal possessive pronouns followed by a noun, particularly the word "MY."

oai_token-act-pair · gpt-3.5-turbo

references to personal possession

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 occup

-0.65

iser

-0.62

 Intercept

-0.58

 Radiation

-0.57

 Liver

-0.57

 directed

-0.57

liness

-0.57

ard

-0.56

 tens

-0.56

ators

-0.55

POSITIVE LOGITS

MY

4.03

MY

2.71

my

1.73

My

1.35

Myth

1.17

mys

1.12

My

1.11

FY

1.06

OUR

0.97

MX

0.96

Activations Density 0.009%

personal possessive pronouns followed by a noun, particularly the word "MY."

references to personal possession

No Comments

No Known Activations