INDEX

Explanations

references to informational content and metadata in documents

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 34-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_34/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.34.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

contentLoaded

-0.91

 يتيمه

-0.81

 Infórmanos

-0.80

Personendaten

-0.75

yarnpkg

-0.71



-0.70

 Wiktionnaire

-0.69

 للاسماء

-0.69

دانشنامهٔ

-0.65

uxxxx

-0.65

POSITIVE LOGITS

Sci

0.42

 Curr

0.39

 Crit

0.37

 meds

0.37

 TECH

0.37

Sys

0.37

 Prep

0.36

 Neurosci

0.36

 Konf

0.35

 Steph

0.35

Activations Density 1.039%

references to informational content and metadata in documents

No Comments

No Known Activations

references to informational content and metadata in documents

No Comments

No Known Activations