INDEX

Explanations

technical terms and programming-related keywords

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

/on/config

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



-0.49

MLLoader

-0.47

tvguidetime

-0.45

 disambiguazione

-0.44

 '\\;'

-0.44

boutin

-0.42

 useAppContext

-0.42

енча

-0.42

ViewFeatures

-0.41

 waer

-0.40

POSITIVE LOGITS

 femeninos

0.44

 expérimentés

0.41

tagPool

0.40

 ordinaires

0.39

 kautta

0.39

tagHelperRunner

0.38

 Brot

0.38



0.38

ConfigureAwait

0.38

ToAdd

0.38

Activations Density 0.000%

technical terms and programming-related keywords

/on/config

No Comments

No Known Activations