INDEX

Explanations

terms related to decision-making processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.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

 kasarigan

-0.59

RunWith

-0.51



-0.50

TagMode

-0.42

 sabar

-0.41

inander

-0.39

SPECIFIED

-0.39

psa

-0.39

 AssemblyCulture

-0.37

writerow

-0.36

POSITIVE LOGITS

 makers

0.75

makers

0.65

 maker

0.64

 Makers

0.59

maker

0.57

 Maker

0.55

findung

0.52

 tomada

0.52

 taken

0.49

 فريبيس

0.49

Activations Density 0.186%

terms related to decision-making processes

No Comments

No Known Activations