INDEX

Explanations

expressions related to understanding and interpreting complex emotions or situations

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

unusual or incomprehensible

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

Efq

-0.49

}]

-0.47

'},

-0.45

 myſelf

-0.44

")]

-0.41

lorette

-0.41

ChildScrollView

-0.40

]');

-0.40

`;

-0.39

ейчас

-0.38

POSITIVE LOGITS

ंदीखरीदारी

0.52

 EconPapers

0.50

 <<<<<<<<<<<<<<

0.46

tat

0.43

tegen

0.41

 outsiders

0.41

횟

0.40

XmlAccessType

0.40

 leef

0.40

nal

0.39

Activations Density 0.043%

expressions related to understanding and interpreting complex emotions or situations

unusual or incomprehensible

No Comments

No Known Activations