INDEX

Explanations

phrases that indicate evaluation, recognition, or definition, particularly related to articles or studies in a formal context

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

 bénévol

-0.39

 évêque

-0.37

 normaux

-0.35

 franqu

-0.35

 étr

-0.35

 Recursos

-0.34

Jegyzetek

-0.34

 internetowa

-0.33

 genética

-0.33

 témoin

-0.33

POSITIVE LOGITS

LookAnd

0.72

 emailAlready

0.55

 ujednoznacz

0.54

thâu

0.54

onViewCreated

0.53

 ostavi

0.49

клопе

0.48

msgTypes

0.48

RenderAtEndOf

0.47

vrolet

0.46

Activations Density 0.811%

phrases that indicate evaluation, recognition, or definition, particularly related to articles or studies in a formal context

No Comments

No Known Activations