INDEX

Explanations

expressions indicating advice or opinions

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

 coö

-1.34

 suspic

-1.31

 emphat

-1.31

 increa

-1.26

 reluct

-1.26

 wherea

-1.24

 effe

-1.23

 embra

-1.20

 uninten

-1.20

 disagre

-1.19

POSITIVE LOGITS

unless

0.80

 unless

0.74

Cannot

0.73

nor

0.71

 Therefore

0.71

cannot

0.69

nor

0.66

must

0.65

 cannot

0.64

Therefore

0.63

Activations Density 0.471%

expressions indicating advice or opinions

No Comments

No Known Activations