INDEX

Explanations

technical and programming-related code structures and commands

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

code and parameters

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

styleType

-0.76

 للمعارف

-0.66

 referenties

-0.57

 pleaſure

-0.56

 ſche

-0.56

 purpoſe

-0.54

 Wikiseite

-0.54

 wikipagina

-0.52

 ſind

-0.51

posedge

-0.50

POSITIVE LOGITS

 бывают

0.40

 бывает

0.39

だそうです

0.38

 Gavin

0.38

 Rowley

0.35

}}^{\

0.35

Gavin

0.35

 считается

0.35

COW

0.34

 Remember

0.34

Activations Density 0.097%

technical and programming-related code structures and commands

code and parameters

No Comments

No Known Activations