INDEX

Explanations

instances of high activation values in sequences or programming contexts

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

Misspelled or non-English "eight"

np_acts-logits-general · gemini-2.0-flash

eight letter words

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

SequentialGroup

-0.57

man

-0.54

 Fourth

-0.54

 NSCoder

-0.53

NCF

-0.52

 Brent

-0.52

Brent

-0.52

 fourth

-0.50

第四

-0.49

AxisAlignment

-0.49

POSITIVE LOGITS

0.89

ate

0.73

 disambiguazione

0.71

 eight

0.59

 Eight

0.57

Ate

0.56

atever

0.56

Notae

0.55

 Acht

0.55

Activations Density 0.040%

instances of high activation values in sequences or programming contexts

Misspelled or non-English "eight"

eight letter words

No Comments

No Known Activations