INDEX

Explanations

phrases related to specific locations and contextual details

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

first occurrences or rarity

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

httphttps

-0.52

省市镇

-0.50

点此举报

-0.47

évaluateur

-0.47

tagHelperRunner

-0.47

InjectAttribute

-0.45

 '\\;'

-0.44

ThemeOverlay

-0.43

 PagesJaunes

-0.42

 disambiguazione

-0.41

POSITIVE LOGITS

 rarest

0.46

Ever

0.42

 Ever

0.40

 unprecedented

0.40

 perdana

0.40

 first

0.39

 Unique

0.39

 groundbreaking

0.38

 pertama

0.37

EVER

0.36

Activations Density 0.037%

phrases related to specific locations and contextual details

first occurrences or rarity

No Comments

No Known Activations