INDEX

Explanations

references to practices and methods in various contexts

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

thschild

-0.56

 pecuni

-0.56

 unterkunft

-0.48

cdb

-0.48

 Decoder

-0.48

 Neub

-0.48

 Anſ

-0.47

 Coulomb

-0.47

 unſ

-0.47

 uranium

-0.47

POSITIVE LOGITS

 PRACTICE

1.01

practice

0.99

 Practice

0.99

Practice

0.98

 practice

0.97

 practise

0.96

PRACTICE

0.93

practi

0.88

 practices

0.82

Practi

0.81

Activations Density 0.138%

references to practices and methods in various contexts

No Comments

No Known Activations