INDEX

Explanations

terms related to references and citations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_7/width_16k/average_l0_92

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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

 فريبيس

-0.47

 TestBed

-0.43

awaiter

-0.42

期刊论文

-0.40

angerschaft

-0.39

 promise

-0.38

verifyException

-0.36

OrBuilder

-0.36

 anger

-0.36

 promised

-0.36

POSITIVE LOGITS

FYI

0.60

Vidite

0.54

FYI

0.52

 Flasche

0.44

FWIW

0.44

 convenience

0.44

 Italij

0.43

ちなみに

0.43

Reminder

0.43

 perusal

0.42

Activations Density 0.271%

terms related to references and citations

No Comments

No Known Activations