INDEX

Explanations

references to "twin" in various contexts

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

twin compounds and phrases

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

 المعيارى

-0.74

忑

-0.68

ophagus

-0.61

efeller

-0.60

URLException

-0.59

 Meksiku

-0.58

Kaynakça

-0.56

 onPage

-0.56

 تانيه

-0.54

chitis

-0.54

POSITIVE LOGITS

 twin

1.17

Twin

1.14

 Twin

1.14

 twins

1.05

twin

1.02

 Twins

0.89

Twins

0.77

 geme

0.63

ツイン

0.59

 double

0.54

Activations Density 0.009%

references to "twin" in various contexts

twin compounds and phrases

No Comments

No Known Activations