INDEX

Explanations

references to evidence and conclusions drawn from examinations or reviews

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

atguigu

-0.52

BagConstraints

-0.50

abetes

-0.49

 Transient

-0.45

 consul

-0.44

cagon

-0.43

 Conservatory

-0.43

MockMvc

-0.42

 europa

-0.42

 Holo

-0.42

POSITIVE LOGITS

eseorang

0.52

 Unterscheidung

0.45

 idéia

0.44

 Anreg

0.43

sproz

0.42

ScopeManager

0.41

TagHelper

0.41

 Züge

0.40

 seeming

0.40

 Domínguez

0.40

Activations Density 0.650%

references to evidence and conclusions drawn from examinations or reviews

No Comments

No Known Activations