INDEX

Explanations

references to software libraries and their functionalities

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

libraries targeting

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

ValueGenerated

-0.43

twimg

-0.43

won

-0.43

 final

-0.36

pagestyle

-0.36

xlabel

-0.36

Eind

-0.35

dataclass

-0.34

ervazione

-0.34

 immediate

-0.34

POSITIVE LOGITS

 Normdatei

0.65

CppMethod

0.63

 libraries

0.63

 library

0.61

libraries

0.59

 biblioteca

0.55

 libs

0.55

 Libraries

0.54

Libraries

0.54

 bibliothèque

0.53

Activations Density 0.031%

references to software libraries and their functionalities

libraries targeting

No Comments

No Known Activations