INDEX

Explanations

programming terms related to intentional design and behavior

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

parsedMessage

-0.62

 ſche

-0.56

 Houſe

-0.55

 houſe

-0.54

ſelf

-0.53

enfone

-0.53

 Chriftian

-0.51

 ſtate

-0.51

NameInMap

-0.51

:✨

-0.48

POSITIVE LOGITS

 inconsistent

0.54

 inconsistency

0.52

 incorrect

0.47

 Unterscheidung

0.47

 Unterschiede

0.45

 inconsistencies

0.44

 erroneous

0.43

 tradição

0.42

 costumbre

0.42

 mistake

0.42

Activations Density 1.201%

programming terms related to intentional design and behavior

No Comments

No Known Activations