INDEX

Explanations

variations of the words "kind" and "sort," indicating a focus on categorization or classification

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

“

-0.57

 Warne

-0.55

 ModelExpression

-0.54

ErrorListener

-0.52

icin

-0.52

 "../../

-0.51

Ce

-0.51

dahl

-0.50

 "../../../

-0.50

 Keys

-0.49

POSITIVE LOGITS

KommentareTeilen

0.94

homonymie

0.92

 BoxFit

0.86

 ſtate

0.86

 purpoſe

0.84

 løpet

0.84

of

0.82

dientemente

0.81

 فريبيس

0.80

katapos

0.79

Activations Density 0.388%

variations of the words "kind" and "sort," indicating a focus on categorization or classification

No Comments

No Known Activations