INDEX

Explanations

phrases indicating comparisons of magnitude or quantity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

kháu

-0.45

ษา

-0.44

Foam

-0.41

加減

-0.41

 initComponents

-0.41

 نش

-0.40

SequentialGroup

-0.40

gms

-0.40

Vest

-0.39

 Finish

-0.39

POSITIVE LOGITS

раздо

0.64

far

0.60

 nahilalakip

0.57

 jauh

0.56

 better

0.54

far

0.53

 BitmapFactory

0.51

 different

0.51

__(/*!

0.50

 dissimilar

0.47

Activations Density 0.290%

phrases indicating comparisons of magnitude or quantity

No Comments

No Known Activations