INDEX

Explanations

numerical data or references in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 surla

-0.55

Rhestr

-0.50

 estadounid

-0.49

WireFormatLite

-0.49

PerformLayout

-0.48

 NSCoder

-0.47

erapeutics

-0.47

MessageOf

-0.47

 betweenstory

-0.46

iſche

-0.46

POSITIVE LOGITS

 يتيمه

0.60

 thereafter

0.35

 كومونز

0.35

 conversely

0.34

吋

0.33

хьтан

0.33

IgnoreCase

0.33

 subsequent

0.32

 navideños

0.32

tæ

0.32

Activations Density 0.413%

numerical data or references in the text

No Comments

No Known Activations