INDEX

Explanations

elements related to movie statistics and categorization

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

watchlist or category labels

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

tonode

-0.65

AnchorTagHelper

-0.51

webElementXpaths

-0.50

 يتيمه

-0.49

 Италијани

-0.49

setVerticalGroup

-0.47

adaptiveStyles

-0.47

Personendaten

-0.46

didSet

-0.45

enterOuterAlt

-0.45

POSITIVE LOGITS

 humeur

0.52

 Kleider

0.48

pulseira

0.47

quivado

0.45

sapato

0.44

न्दर्भ

0.43

 Odour

0.41

andExpect

0.41

Caractéristiques

0.41

Senha

0.40

Activations Density 0.024%

elements related to movie statistics and categorization

watchlist or category labels

No Comments

No Known Activations