INDEX

Explanations

expressions related to personal beliefs or experiences

oai_token-act-pair · gpt-3.5-turbo

personal reflections on life experiences

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

ãĥĥãĥĪ

-0.65

viks

-0.64

adelphia

-0.62

ackers

-0.62

ãĥī

-0.62

 Officials

-0.62

Pwr

-0.61

ItemTracker

-0.61

ranch

-0.60

©¶æ¥µ

-0.59

POSITIVE LOGITS

my

2.06

 myself

1.68

1.66

me

1.43

MY

1.36

 mine

1.34

my

1.34

My

1.33

My

1.32

1.21

Activations Density 1.128%

expressions related to personal beliefs or experiences

personal reflections on life experiences

No Comments

No Known Activations