INDEX

Explanations

references to polls and polling-related terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_0/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.0.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

 alimentaire

-0.37

 HttpNotFound

-0.37

 Gubernur

-0.37

 Ambiental

-0.36

 Económica

-0.35

 peindre

-0.34

Rs

-0.34

 Constitucional

-0.33

 Moscú

-0.33

 constitución

-0.33

POSITIVE LOGITS

 penetrated

0.66

 poll

0.66

opt

0.66

 penetrates

0.66

AndroidJUnit

0.64

penet

0.64

 penet

0.64

Pen

0.63

Pen

0.62

 penetrate

0.61

Activations Density 0.182%

references to polls and polling-related terminology

No Comments

No Known Activations