INDEX

Explanations

programming terminology and constructs related to data handling and decoding

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

Decoding data using `decoder` and `decodeSingular` methods

np_acts-logits-general · gemini-2.0-flash

me great

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 30-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_30/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.30.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

PreferredItem

-0.55

alyptus

-0.54

 wikipagina

-0.52

EDEFAULT

-0.51

Datuak

-0.50

toMatchSnapshot

-0.49

 strdup

-0.49

momix

-0.49

 stanga

-0.49

ftagPool

-0.47

POSITIVE LOGITS

0.81

(&

0.78

0.70

,&

0.61

=&

0.57

:&

0.56

\&

0.56

(&$

0.55

&$

0.55

ref

0.54

Activations Density 0.242%

programming terminology and constructs related to data handling and decoding

Decoding data using `decoder` and `decodeSingular` methods

me great

No Comments

No Known Activations