INDEX

Explanations

mention of health insurance related terms and policies, particularly around enrollment and coverage

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

 encomp

-1.51

 desir

-1.50

 depic

-1.49

 effe

-1.41

 guarante

-1.39

 inev

-1.39

 perfet

-1.38

 suscep

-1.37

 eyel

-1.37

 unden

-1.37

POSITIVE LOGITS

 despite

0.70

 themselves

0.70

 regardless

0.63

 while

0.61

 because

0.61

 either

0.59

 unless

0.58

due

0.58

Their

0.58

 during

0.57

Activations Density 0.685%

mention of health insurance related terms and policies, particularly around enrollment and coverage

No Comments

No Known Activations