INDEX

Explanations

expressions and actions related to pleasing and comfort

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

WithIdentifier

-0.59

endphp

-0.59

NewUrlParser

-0.56

Inspiration

-0.56

antwortung

-0.55

 celo

-0.55

MemoryWarning

-0.53

 страх

-0.53

viewDidLoad

-0.53

gew

-0.52

POSITIVE LOGITS

Pleasant

1.02

 agradável

1.00

 pleasing

0.97

 Pleasant

0.97

 pleasant

0.96

 agradable

0.88

 agrad

0.87

pleasant

0.84

 agréable

0.82

 прият

0.80

Activations Density 0.367%

expressions and actions related to pleasing and comfort

No Comments

No Known Activations