INDEX

Explanations

terms related to hierarchical structures or elements in a technical context

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-mlp/layer_6/width_131k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.6.hook_mlp_out

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

########.

-0.65

contentLoaded

-0.46

kháu

-0.42

 faudrait

-0.42

sizeCache

-0.41

:][

-0.41

енча

-0.40

SequentialGroup

-0.40

ReusableCell

-0.39

 stanovnika

-0.39

POSITIVE LOGITS

Chham

0.39

 Skal

0.38

 defaultstate

0.37

oury

0.36

cina

0.36

transQ

0.36

 intermedio

0.35

 quebra

0.35

 κό

0.35

Ɓ

0.35

Activations Density 0.025%

terms related to hierarchical structures or elements in a technical context

No Comments

No Known Activations