INDEX

Explanations

statements about controversial statements or actions

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

 gracilis

-0.75

 zima

-0.69

 serai

-0.67

Fuckin

-0.66

 moze

-0.63

 kawaida

-0.60

 alpes

-0.60

 occidentalis

-0.58

vne

-0.57

 rano

-0.57

POSITIVE LOGITS

 Needless

0.55

Needless

0.52

 Edizioni

0.50

 paradiso

0.49

 dieux

0.48

 Thankfully

0.48

 Cinéma

0.48

InjectAttribute

0.47

toBeDefined

0.47

 natale

0.47

Activations Density 1.610%

statements about controversial statements or actions

No Comments

No Known Activations