INDEX

Explanations

terms related to feature extraction and evaluation methods in convolutional neural networks (CNNs)

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

discriminative learning

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

aarrggbb

-0.46

RegressionTest

-0.42

TagHelper

-0.36

 inaczej

-0.36

where

-0.35

|}{\

-0.35

autant

-0.34

 cadeaux

-0.34

 resourceCulture

-0.32

 gethan

-0.32

POSITIVE LOGITS

 kasarigan

0.63

TagMode

0.59

بوابة

0.52

/**

0.52

 autorytatywna

0.50

PYX

0.46

TestingModule

0.45

OGND

0.44

 snippetHide

0.44

❋

0.44

Activations Density 0.205%

terms related to feature extraction and evaluation methods in convolutional neural networks (CNNs)

discriminative learning

No Comments

No Known Activations