INDEX

Explanations

references to the concept of "con" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

inafter

-0.68

ISTICS

-0.64

IsContent

-0.57

 BoxFit

-0.56

 CallOverrides

-0.55

unsplash

-0.55

abestanden

-0.53

rootDir

-0.53

Wikis

-0.53

Reprodução

-0.53

POSITIVE LOGITS

Con

0.76

Con

0.76

CON

0.71

con

0.69

CON

0.59

 autorytatywna

0.56

con

0.56

 Conquest

0.53

gink

0.51

 cones

0.51

Activations Density 0.120%

references to the concept of "con" in various contexts

No Comments

No Known Activations