INDEX

Explanations

references to authorship and contributions in studies

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

 MonoBehaviour

-0.45

 herido

-0.35

fur

-0.34

GetAsync

-0.33

 Савезне

-0.33

Nội

-0.33

 fantasmas

-0.32

μφ

-0.31

 ··

-0.31

 insuffisamment

-0.30

POSITIVE LOGITS

SequentialGroup

0.63

 насељу

0.54

 <<<<<<<<<<<<<<

0.52

abstractmethod

0.49

XmlAccessType

0.46

 Competition

0.46

'{@

0.46

 Drawing

0.45

UnsafeEnabled

0.45

‌آم

0.45

Activations Density 0.020%

references to authorship and contributions in studies

No Comments

No Known Activations