INDEX

Explanations

technical details and mechanisms related to scientific explanations or theories

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

DockStyle

-0.59

Jeografia

-0.50

oredCriteria

-0.49

----</

-0.48

potranspiration

-0.47

EndContext

-0.47

 fantaisie

-0.46

awaiter

-0.46

ंदीखरीदारी

-0.45

 EClass

-0.45

POSITIVE LOGITS

 interplay

0.40

 explicar

0.40

 penting

0.38

 understanding

0.37

 Réponses

0.37

 actionMode

0.37

 merit

0.37

 explicación

0.37

migrationBuilder

0.36

interprétation

0.36

Activations Density 0.122%

technical details and mechanisms related to scientific explanations or theories

No Comments

No Known Activations