INDEX

Explanations

references to the concept of "wholeness" or "entirety."

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

Precedes words indicating scope or extent

np_acts-logits-general · gemini-2.0-flash

totality and scope

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

 bacilli

-0.56

 massima

-0.56

 nakalista

-0.53

 alguna

-0.50

 alguno

-0.49

 creș

-0.48

 poppies

-0.46

 nenhuma

-0.46

 algún

-0.46

 piety

-0.46

POSITIVE LOGITS

 thing

1.13

 entire

0.99

 gamut

0.92

 length

0.82

 process

0.81

 contents

0.80

 duration

0.80

 whole

0.76

 THING

0.75

 spectrum

0.73

Activations Density 0.073%

references to the concept of "wholeness" or "entirety."

Precedes words indicating scope or extent

totality and scope

No Comments

No Known Activations