INDEX

Explanations

references to user reviews and feedback mechanisms

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

review and rating prompts

np_acts-logits-general · gemini-2.5-flash-lite

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

transQ

-0.61

ſelf

-0.52

 Majefty

-0.51

afu

-0.50

 fubject

-0.48

 surla

-0.47

 الرياضيه

-0.47

protoimpl

-0.47

 durs

-0.46

weck

-0.45

POSITIVE LOGITS

 ligiloj

0.38

stüt

0.37

 Encu

0.36

 Aeropuerto

0.34

 compito

0.33

 Wikiseite

0.33

 époux

0.32

 ambassade

0.32

erschein

0.32

 instalação

0.32

Activations Density 0.004%

references to user reviews and feedback mechanisms

review and rating prompts

No Comments

No Known Activations