INDEX

Explanations

terms related to clustering and unsupervised learning methods

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

clustering and unsupervised methods

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

 kılı

-0.43

 safe

-0.39

 aikana

-0.38

раздо

-0.38

 einger

-0.37

Гаран

-0.37

 raiſ

-0.37

 légal

-0.37

Dressing

-0.36

Dine

-0.36

POSITIVE LOGITS

 ProtoMessage

0.71

kmeans

0.68

 clustering

0.57

Tikang

0.56

期刊论文

0.56

 OMITBAD

0.53

 Clustering

0.52

ThroughAttribute

0.50

entification

0.50

InputBorder

0.50

Activations Density 0.026%

terms related to clustering and unsupervised learning methods

clustering and unsupervised methods

No Comments

No Known Activations