INDEX

Explanations

elements related to complexity and detail in descriptions or narratives

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

but only

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

 arşivlendi

-0.47

 ThemeData

-0.44

っこう

-0.43

 surla

-0.41

flashdata

-0.40

Geplaatst

-0.40

 crossorigin

-0.39

 Paglinawan

-0.38

RegistryLite

-0.38

 ویکی‌پدیای

-0.37

POSITIVE LOGITS

 CreateTagHelper

0.53

ⓧ

0.52

Personendaten

0.51

 snippetHide

0.46

########.

0.45

 rather

0.45

 oprot

0.44

 piuttosto

0.41

UMBIA

0.41

 <=",

0.41

Activations Density 0.081%

elements related to complexity and detail in descriptions or narratives

but only

No Comments

No Known Activations