INDEX

Explanations

requests or expressions of intent

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 34-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_34/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.34.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

탤

-0.47

HideFlags

-0.42

SerializedSize

-0.41

 talks

-0.39

inSlope

-0.39

chemise

-0.39

 Discus

-0.38

GOTREF

-0.38

 discussions

-0.36

Discus

-0.36

POSITIVE LOGITS

 congratulate

0.60

 отметить

0.57

 reiterate

0.55

 clarify

0.55

 mention

0.55

 remark

0.54

say

0.53

 apologize

0.52

 aclarar

0.51

 preface

0.51

Activations Density 0.311%

requests or expressions of intent

No Comments

No Known Activations