INDEX

Explanations

sequences that denote data structures or code elements in programming contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 25-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_25/width_16k/average_l0_116

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.25.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

<bos>

-0.56

 allAfrica

-0.17

ρυ

-0.17

>{@

-0.17

λοι

-0.17

 VIDEOTAPE

-0.16

 queſta

-0.16

gettime

-0.15

 vägen

-0.14

xffffffff

-0.14

POSITIVE LOGITS

ⓧ

0.49

ViewImports

0.30

 betweenstory

0.30

 nahilalakip

0.29

الدراسه

0.26

SpringBootTest

0.24

новништво

0.24

География

0.24

 surla

0.24

Égypte

0.23

Activations Density 31.924%

sequences that denote data structures or code elements in programming contexts

No Comments

No Known Activations