INDEX

Explanations

phrases related to admiration for a particular character

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Configuration

jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

standard

Context Size

1,024

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 utop

-1.26

 makro

-1.26

 dises

-1.23

 territo

-1.22

 meis

-1.20

 kram

-1.19

 tomat

-1.17

 ananas

-1.16

 lele

-1.15

 franz

-1.12

POSITIVE LOGITS

 possesses

0.83

is

0.80

 excels

0.78

 possess

0.77

 deserves

0.76

has

0.74

 behaves

0.73

0.72

’

0.72

 embodies

0.70

Activations Density 0.506%

phrases related to admiration for a particular character

No Comments

No Known Activations