INDEX

Explanations

patterns and structures in markup language or code

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

Abbreviations followed by left bit shift

np_acts-logits-general · gemini-2.0-flash

multimauDDmicropolyijer

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

MigrationBuilder

-1.37

 betweenstory

-1.27

parsedMessage

-1.22

rungsseite

-1.19

SharedDtor

-1.18

Diwedd

-1.18

principalColumn

-1.17

 мәкал

-1.16

 nakalista

-1.16

AndEndTag

-1.14

POSITIVE LOGITS

0.48

0.45

</h2>

0.42

</strong>

0.41

</em>

0.41

0.40

<unused63>

0.40

<unused61>

0.39

<unused62>

0.39

0.38

Activations Density 0.123%

patterns and structures in markup language or code

Abbreviations followed by left bit shift

multimauDDmicropolyijer

No Comments

No Known Activations