INDEX

Explanations

emotional states and perceptions about individuals or situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 27-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_27/width_16k/average_l0_118

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.27.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

InputBorder

-0.48

DeleteBehavior

-0.44

XmlAccessType

-0.43

ConstraintMaker

-0.43

LookAnd

-0.42

setVerticalGroup

-0.40

ArgsConstructor

-0.38

WebElementEntity

-0.37

 AssemblyTitle

-0.36

odly

-0.36

POSITIVE LOGITS

 intention

0.61

つもり

0.60

 intended

0.60

 intent

0.59

 intentions

0.57

intended

0.56

intention

0.56

本当は

0.56

 secretly

0.56

planned

0.55

Activations Density 0.636%

emotional states and perceptions about individuals or situations

No Comments

No Known Activations