INDEX

Explanations

phrases related to evaluation and perceptions of importance

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

sections and important factors

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

gridy

-0.41

 désolés

-0.38

fromnode

-0.38

ритори

-0.37

 Rides

-0.37

 colspan

-0.35

➋

-0.35

الدراسه

-0.35

UnitTesting

-0.35

 ролях

-0.34

POSITIVE LOGITS

ViewFeatures

0.56

NameInMap

0.48

 oprot

0.47

Autoritní

0.44

mergeFrom

0.44

المكان

0.43

Romain

0.42

 internetowa

0.41

 humaine

0.41

->___

0.41

Activations Density 0.136%

phrases related to evaluation and perceptions of importance

sections and important factors

No Comments

No Known Activations