INDEX

Explanations

punctuations and symbols used in programming or structured data

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 graphique

-0.50

 سكانية

-0.46

 seguimiento

-0.44

 apprécier

-0.44

 refuerzo

-0.44

 becomes

-0.43

anderen

-0.43

 colectiva

-0.43

 recueil

-0.42

 intercambio

-0.42

POSITIVE LOGITS

 zero

0.71

 Zero

0.67

 ZERO

0.67

zero

0.64

Zero

0.63

zeros

0.61

__((

0.60

 zeros

0.58

 zeroes

0.58

ゼロ

0.54

Activations Density 0.208%

punctuations and symbols used in programming or structured data

No Comments

No Known Activations