INDEX

Explanations

references to objectives or targets

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_1/width_16k/average_l0_69

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

 autop

-0.50

+:+

-0.49



-0.49

 brid

-0.47

 unic

-0.47

 विश्वसनीयता

-0.47

 purch

-0.46

 nucle

-0.46

 predis

-0.46

 pend

-0.45

POSITIVE LOGITS

 goals

1.60

Goals

1.38

goals

1.35

 goal

1.34

goal

1.29

 aims

1.23

 Goals

1.23

 objectives

1.22

 Ziele

1.21

 GOALS

1.20

Activations Density 0.159%

references to objectives or targets

No Comments

No Known Activations