INDEX

Explanations

elements related to parameters, settings, and adjustments in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ADORA

-0.34

 compétence

-0.33

sum

-0.32

isRequired

-0.31

fatter

-0.31

over

-0.31

Allora

-0.31

Ken

-0.30

lunda

-0.30

service

-0.30

POSITIVE LOGITS

########.

0.79

ftagPool

0.74

msgTypes

0.71

ViewFeatures

0.68

QMetaType

0.63

GOTREF

0.60

 informée

0.60

 <=",

0.59

 Bioaccumulative

0.57

PyExc

0.56

Activations Density 0.856%

elements related to parameters, settings, and adjustments in various contexts

No Comments

No Known Activations