INDEX

Explanations

parentheses in various contexts

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

acronyms and abbreviations

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

 happy

-0.33

 very

-0.30

 short

-0.29

 Italian

-0.29

 Exteriores

-0.28

 give

-0.28

 visitors

-0.28

 extranjera

-0.28

old

-0.28

 fresh

-0.28

POSITIVE LOGITS

Personendaten

0.84

="@+

0.79

 يتيمه

0.79

 noDo

0.74

 kasarigan

0.74

:✨

0.74

 Paglinawan

0.72

ValueStyle

0.72

تقاوى

0.71

OGND

0.70

Activations Density 0.111%

parentheses in various contexts

acronyms and abbreviations

No Comments

No Known Activations