INDEX

Explanations

several variations of the word "in" as well as instances of conditions or negations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_11/width_16k/average_l0_80

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.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

;'>

-0.62

 Ceramby

-0.61

 SignUp

-0.60

-//

-0.58

—

-0.58

TagHelper

-0.57

 Europ

-0.57

blurRadius

-0.57

Abb

-0.56

fighting

-0.56

POSITIVE LOGITS

 متعلقه

0.71

MigrationBuilder

0.67

 propOrder

0.67

شهاد

0.67

حوالہ

0.66

unhofer

0.60

 magát

0.59

telar

0.59

Misty

0.59

 באתר

0.59

Activations Density 0.076%

several variations of the word "in" as well as instances of conditions or negations

No Comments

No Known Activations