INDEX

Explanations

phrases indicating agreement or consistency within various studies and results

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_22/width_16k/average_l0_123

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

 Walkover

-0.53

},{

-0.46

 transfieras

-0.38

 tables

-0.37

脱

-0.37

jok

-0.36

 noDo

-0.35

 trời

-0.35

énario

-0.35

 WHOLE

-0.35

POSITIVE LOGITS

 conferma

0.66

 confirms

0.65

 confirming

0.62

 confirm

0.60

 confirmó

0.59

 predicted

0.56

 confirmación

0.56

 expected

0.56

 confirmar

0.56

 potwier

0.54

Activations Density 1.092%

phrases indicating agreement or consistency within various studies and results

No Comments

No Known Activations