INDEX

Explanations

references to quantifiable measurements or numerical values

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

CreateIndex

-0.34

行

-0.32

 liste

-0.31

ro

-0.31

 lists

-0.31

ar

-0.30

 nothing

-0.29

 ре

-0.29

ành

-0.29

</h3>

-0.28

POSITIVE LOGITS

 يتيمه

0.77

MLLoader

0.66

нгред

0.66

 '\\;'

0.61

SBATCH

0.59

 contextLoads

0.56

 otomatig

0.56

fjspx

0.55

 ویکی‌پدی

0.54

GraphicsUnit

0.54

Activations Density 0.238%

references to quantifiable measurements or numerical values

No Comments

No Known Activations