INDEX

Explanations

terms related to image processing and transformation techniques

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

transforms and protocols

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

AddTagHelper

-0.94

 ujednoznacz

-0.73

rungsseite

-0.69

 Normdatei

-0.67

SBATCH

-0.67

 Italijanski

-0.65

expandindo

-0.64

点此举报

-0.64

#+#

-0.62

 Taktlose

-0.62

POSITIVE LOGITS

transforms

0.52

 transforms

0.40

Transforms

0.40

 transforma

0.39

ToTensor

0.39

 routines

0.37

 routine

0.36

 transform

0.34

RectangleBorder

0.34

 transformations

0.33

Activations Density 0.001%

terms related to image processing and transformation techniques

transforms and protocols

No Comments

No Known Activations