INDEX

Explanations

processes and documentation related to research and study evaluations

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

execSQL

-0.34

ool

-0.31

ше

-0.29

Kjelder

-0.28

Host

-0.28

yen

-0.28

 Host

-0.27

 तह

-0.27

込め

-0.27

hide

-0.26

POSITIVE LOGITS

 correctness

0.67

GIVEREF

0.66

httphttps

0.65

 accuracy

0.65

 incorrect

0.63

WireFormatLite

0.61

 inaccuracies

0.61

 inaccurate

0.60

 Incorrect

0.60

oredCriteria

0.60

Activations Density 0.621%

processes and documentation related to research and study evaluations

No Comments

No Known Activations