INDEX

Explanations

phrases related to confusion and uncertainty

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

switch

-0.35

Kontra

-0.34

 encima

-0.33

chyma

-0.33

 contraire

-0.32

ズル

-0.31

 Administrativna

-0.29

TITUDE

-0.29

 együtt

-0.29

 מל

-0.29

POSITIVE LOGITS

 betweenstory

0.66

 incoherent

0.65

 lunatic

0.65

 delusional

0.62

 delirium

0.60

coher

0.58

 crazy

0.56

 irrational

0.56

 disordered

0.54

UnitTesting

0.54

Activations Density 0.100%

phrases related to confusion and uncertainty

No Comments

No Known Activations