INDEX

Explanations

mentions of formal documentation or references

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

foreign words / specific identifiers

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

fører

-0.38

Bla

-0.33



-0.33

standes

-0.33

 dolayı

-0.33

 pomocy

-0.32

 أعلام

-0.32

 cobrar

-0.32

 étroite

-0.31

ÍST

-0.31

POSITIVE LOGITS

Tikang

0.70

 Мексичка

0.68

Hochspringen

0.65

تقاوى

0.64

 surla

0.61

 للاسماء

0.60

 PagesJaunes

0.60

 ویکی‌پدی

0.59

:✨

0.59

 nakalista

0.59

Activations Density 0.000%

mentions of formal documentation or references

foreign words / specific identifiers

No Comments

No Known Activations