INDEX

Explanations

terms related to applications and application contexts within a programming or technical framework

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

loyees

-0.39

words

-0.34

ChildScrollView

-0.32

sal

-0.32

setz

-0.32

ANGO

-0.32

 propOrder

-0.30

夠

-0.30

apache

-0.30

dafx

-0.30

POSITIVE LOGITS

程式

0.55

 saveiro

0.55

AutoScaleMode

0.54

 rhestr

0.52

ніципалі

0.51

likation

0.51

 للمعارف

0.50

 myſelf

0.50

ècie

0.49

 transfieras

0.47

Activations Density 0.211%

terms related to applications and application contexts within a programming or technical framework

No Comments

No Known Activations

terms related to applications and application contexts within a programming or technical framework

No Comments

No Known Activations