INDEX

Explanations

syntax elements and structures used in programming or markup languages

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

programming symbols and keywords

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

httphttps

-0.63

сылкі

-0.62

CodeAttribute

-0.62

日閲覧

-0.57

 مشين

-0.55

colhead

-0.54

OrWhiteSpace

-0.51

 arşivlendi

-0.50

 tabPage

-0.49

ResponseWriter

-0.49

POSITIVE LOGITS

eseorang

0.42

+#+#

0.40

 beina

0.39

 enamorado

0.38

stridium

0.37

 teil

0.36

 sprend

0.36

 asistir

0.35

 origini

0.35

 presentaciones

0.35

Activations Density 0.040%

syntax elements and structures used in programming or markup languages

programming symbols and keywords

No Comments

No Known Activations