INDEX

Explanations

references to the head and associated terms

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

aarrggbb

-0.62

URLException

-0.51

ErrorException

-0.50

PMI

-0.48

 Pact

-0.48

出版年

-0.47

 Paradox

-0.47

 esporte

-0.46

 Vari

-0.46

oarece

-0.46

POSITIVE LOGITS

Head

1.30

 Head

1.27

 HEAD

1.24

 head

1.24

head

1.23

HEAD

1.13

 heads

1.07

 Heads

1.04

heads

0.98

Heads

0.92

Activations Density 0.178%

references to the head and associated terms

No Comments

No Known Activations