INDEX

Explanations

references to arguments, complaints, or actions pertaining to rights and legal issues

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

followed by "because" or reasons

np_acts-logits-general · gemini-2.0-flash

explaining why

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 argc

-0.63

IContainer

-0.62

EDO

-0.62

argout

-0.62

$_,

-0.60

bzero

-0.59

)}$

-0.58

ays

-0.57

tioners

-0.57

iev

-0.56

POSITIVE LOGITS

 because

1.16

 only

1.03

 mainly

0.97

 primarily

0.95

 uniquement

0.90

 porque

0.86

 потому

0.86

 BECAUSE

0.84

 fordi

0.83

because

0.81

Activations Density 1.263%

references to arguments, complaints, or actions pertaining to rights and legal issues

followed by "because" or reasons

explaining why

No Comments

No Known Activations