INDEX

Explanations

high-frequency numerical values or counts in a programming context

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

Preceding tokens of "before" or "having"

np_acts-logits-general · gemini-2.0-flash

made before, having a

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

 time

-0.79

time

-0.68

Time

-0.67

Personendaten

-0.65

 Time

-0.63

 TIME

-0.53

 ujednoznacz

-0.52

TIME

-0.52

 updateTime

-0.52

 compagnia

-0.51

POSITIVE LOGITS

المناصب

0.61

 JpaRepository

0.53

 становника

0.53

 plotly

0.51

NewUrlParser

0.48

Traceback

0.47

reactstrap

0.47

PreferredItem

0.46

 parlant

0.46

WSER

0.46

Activations Density 0.033%

high-frequency numerical values or counts in a programming context

Preceding tokens of "before" or "having"

made before, having a

No Comments

No Known Activations