INDEX

Explanations

the presence of specific control or intervention keywords in the context of procedures or actions

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

with mixed casing or capitalization

np_acts-logits-general · gemini-2.0-flash

FDA approval

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

setVerticalGroup

-0.69

();)

-0.68

tanleria

-0.66

łaszcza

-0.64

MENAFN

-0.61

^{+}\

-0.61

StructEnd

-0.60

rungsseite

-0.60

++)

-0.60

 protoimpl

-0.59

POSITIVE LOGITS

kit

0.52

 snippetHide

0.49

 Atmos

0.44

廊

0.43

KIT

0.43

囊

0.43

 عنه

0.42

Perché

0.41

씩

0.41

 деле

0.41

Activations Density 0.038%

the presence of specific control or intervention keywords in the context of procedures or actions

with mixed casing or capitalization

FDA approval

No Comments

No Known Activations