INDEX

Explanations

words related to processes and actions within various contexts

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

alignment #

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

Hentet

-0.57

 surla

-0.47

IUrlHelper

-0.47

})`

-0.46

Zaj

-0.45

 localVar

-0.44

errHandler

-0.44

modelBuilder

-0.44

anti

-0.43

bibfield

-0.42

POSITIVE LOGITS

NameInMap

0.44

 réaliste

0.38

Deletes

0.37

 Alquiler

0.36

 trouvez

0.36

betreuung

0.35



0.35

 outdoors

0.34

ึกษา

0.34

iseerd

0.34

Activations Density 0.069%

words related to processes and actions within various contexts

alignment #

No Comments

No Known Activations