INDEX

Explanations

occurrences of the word "<bos>", indicating the start of a new section or content block

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

medical, scientific, or technical terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 16-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_16/width_65k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.16.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

GraphicsUnit

-0.66

 setuptools

-0.56

الغ

-0.54

 exemples

-0.52

eway

-0.51

æa

-0.50

 IndexPath

-0.50

τία

-0.50

 Sigh

-0.49

 στε

-0.49

POSITIVE LOGITS

Datuak

0.76

 ujednoznacz

0.67

__":

0.67

клопе

0.64

 adaptive

0.61

 estimés

0.60

__':

0.58

 gainera

0.58

 OFDb

0.58

databind

0.57

Activations Density 0.254%

occurrences of the word "<bos>", indicating the start of a new section or content block

medical, scientific, or technical terms

No Comments

No Known Activations