INDEX

Explanations

concepts and terms related to saliency maps and feature extraction in image processing

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

saliency map

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

MemoryWarning

-0.53

ValueStyle

-0.38

 AutoMapper

-0.38

SequentialGroup

-0.37

PhysRevD

-0.37

 Municipalidad

-0.35

RTDA

-0.35

uttavia

-0.34

Itoa

-0.34

)_/¯

-0.33

POSITIVE LOGITS

出版年

0.55

 perturb

0.47

 betweenstory

0.46

 perturbation

0.45

ɹ

0.44

 pique

0.42

 STRE

0.42

arsis

0.42

 fevere

0.41

 Toxicol

0.41

Activations Density 0.070%

concepts and terms related to saliency maps and feature extraction in image processing

saliency map

No Comments

No Known Activations