INDEX

Explanations

mathematical notations and symbols used in formal definitions and expressions

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

-0.46

-0.42

is

-0.41

 only

-0.41

 even

-0.39

but

-0.38

 combined

-0.37

 there

-0.37

 always

-0.36

 alone

-0.36

POSITIVE LOGITS

 defaultstate

0.77

 Infórmanos

0.75

$_(

0.74

хьтан

0.73

MLLoader

0.72

 Waray

0.69

LabelTagHelper

0.68

ðsíða

0.68

aarrggbb

0.68

rungsseite

0.68

Activations Density 0.566%

mathematical notations and symbols used in formal definitions and expressions

No Comments

No Known Activations