INDEX

Explanations

occurrences of the word "Happy" and variations of it, as well as the term "plain."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_0/width_65k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.0.hook_mlp_out

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

GraphicsUnit

-0.78

Xna

-0.77

 فريبيس

-0.75

 betweenstory

-0.66

 &___

-0.65

OrNil

-0.65

 XCTest

-0.64

 createState

-0.64

çons

-0.64

"}")

-0.63

POSITIVE LOGITS

Jereo

0.68

 kanya

0.60

BASELINE

0.60

 DBNull

0.59

 løpet

0.57

ymce

0.55

NgModule

0.54

 Dumnezeu

0.53

astronomie

0.53

 Jésus

0.52

Activations Density 0.018%

occurrences of the word "Happy" and variations of it, as well as the term "plain."

No Comments

No Known Activations