INDEX

Explanations

terms related to layers or components in technical or scientific contexts

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

prepared systems and items

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

AnchorStyles

-0.44

 tố

-0.42

 CGRectMake

-0.39

SQLiteDatabase

-0.38

imple

-0.35

nextProps

-0.34

xase

-0.33

pl

-0.33

pegno

-0.32

Negoti

-0.32

POSITIVE LOGITS

 ویکی‌پدی

0.77

TypedDataSet

0.65

 Paglinawan

0.62

 resourceCulture

0.60

 OMITBAD

0.52

 copies

0.51

Copies

0.49

хьтан

0.49

tagHelperRunner

0.48

 OFDb

0.48

Activations Density 0.264%

terms related to layers or components in technical or scientific contexts

prepared systems and items

No Comments

No Known Activations