INDEX

Explanations

sections that emphasize key points or main ideas

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

 pelúcia

-0.61

 maioria

-0.53

 Públicas

-0.50

 sereia

-0.48

 zakup

-0.47

 secado

-0.47

 zagran

-0.47

 współpra

-0.47

 zapach

-0.47

 kerugian

-0.46

POSITIVE LOGITS

point

1.40

Point

1.38

 Point

1.33

POINT

1.28

 POINT

1.26

 point

1.24

poin

1.10

points

1.05

 poin

1.00

 points

0.98

Activations Density 0.144%

sections that emphasize key points or main ideas

No Comments

No Known Activations