INDEX

Explanations

statistical comparisons and measurements related to size, potency, and environmental factors

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

comparative and superlative adjectives

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

 мәкал

-0.71

OGND

-0.69

 utafitiHapana

-0.68

 Normdatei

-0.63

 AttributeSet

-0.63

queryInterface

-0.61

 defaultstate

-0.60

HasAnnotation

-0.59

 beſch

-0.59

IsContent

-0.58

POSITIVE LOGITS

 strongest

0.38

 thermique

0.32

 wealthiest

0.32

 برتر

0.31

 untere

0.31

domés

0.31

createCell

0.30

 telefónica

0.30

 best

0.30

 principaux

0.30

Activations Density 0.075%

statistical comparisons and measurements related to size, potency, and environmental factors

comparative and superlative adjectives

No Comments

No Known Activations