INDEX

Explanations

phrases related to navigation and movement through physical spaces

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

turning or corners

np_acts-logits-general · gemini-2.0-flash

turns and corners

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

#+#

-0.56

aarrggbb

-0.54

writerow

-0.53

 ExecuteAsync

-0.53

 Offisielt

-0.52

viewDidLoad

-0.51

asci

-0.50

 fathoms

-0.50

 Shoemaker

-0.49

mallow

-0.49

POSITIVE LOGITS

 turn

1.73

 turns

1.61

turn

1.58

 corner

1.50

Turn

1.48

 turning

1.47

 Turn

1.44

 corners

1.34

turns

1.33

 TURN

1.33

Activations Density 0.300%

phrases related to navigation and movement through physical spaces

turning or corners

turns and corners

No Comments

No Known Activations