INDEX

Explanations

terms related to unsupervised learning and concepts specific to conductors and their properties

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 2-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

 brioche

-0.43

 consultato

-0.43

setVerticalGroup

-0.43

rój

-0.42

 disponibilités

-0.41

ximadamente

-0.40

 frappé

-0.39

fjspx

-0.39

 مهد

-0.38

omiast

-0.38

POSITIVE LOGITS

 lenker

0.67

Eagle

0.57

eagle

0.56

 Eagle

0.56

 betweenstory

0.54

á

0.50

nexpected

0.49

 smoke

0.49

 misunderstand

0.47

jLabel

0.46

Activations Density 0.284%

terms related to unsupervised learning and concepts specific to conductors and their properties

No Comments

No Known Activations