INDEX

Explanations

references to legal documents or case citations

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

or phrases after periods

np_acts-logits-general · gemini-2.0-flash

sentence endings

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

何より

-0.80

 therefore

-0.79

 furthermore

-0.75

 moreover

-0.75

 hence

-0.73

 ezért

-0.72

therefore

-0.72

Поэтому

-0.72

 Derfor

-0.71

 Apalagi

-0.70

POSITIVE LOGITS

 darin

0.94

 Allí

0.93

 Briefly

0.92

 therein

0.82

 wherein

0.69

そちら

0.69

在这个

0.69

 فيه

0.68

Titled

0.68

裡面

0.65

Activations Density 0.516%

references to legal documents or case citations

or phrases after periods

sentence endings

No Comments

No Known Activations