INDEX

Explanations

percentage values and numerical data associated with varying formats and categories

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_11/width_16k/average_l0_80

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.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

 تكبرها

-0.71

 Ludlow

-0.69

دانشنامهٔ

-0.64

 Meksiku

-0.64

 '\\;'

-0.62

 Италијани

-0.62

rrrrrrrr

-0.61

skjaer

-0.61

 parche

-0.60

 Idris

-0.59

POSITIVE LOGITS

enumi

0.90

 nélk

0.66

 وتسجيلات

0.63

lüğ

0.60

\%

0.59

\%

0.59

URIComponent

0.58

󠁢

0.57

räck

0.57

]='\

0.56

Activations Density 0.091%

percentage values and numerical data associated with varying formats and categories

No Comments

No Known Activations