INDEX

Explanations

punctuation marks, specifically full stops

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

user followed by code

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

aarrggbb

-0.69

enumi

-0.68

verifyException

-0.65

 betweenstory

-0.65

 queſta

-0.63

MENAFN

-0.62

 disponibilités

-0.58

 HasFactory

-0.58

 يتيمه

-0.58

enumii

-0.56

POSITIVE LOGITS

*/),

0.43

 autorytatywna

0.38

 aislada

0.34

])).

0.33

*/)

0.32

=").

0.31

 casada

0.31

 rotina

0.31

 murni

0.30

 líquida

0.29

Activations Density 0.000%

punctuation marks, specifically full stops

user followed by code

No Comments

No Known Activations