INDEX

Explanations

expressions related to the concept of "best" or optimal choices

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 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

 coroa

-0.59

 bandeira

-0.56

 camiset

-0.54

colare

-0.49

SequentialGroup

-0.47

Filmografia

-0.47

 folha

-0.45

 McLeod

-0.45

 Carrillo

-0.45

 atos

-0.45

POSITIVE LOGITS

Better

1.01

 Better

0.96

Best

0.92

better

0.91

 BETTER

0.90

 BEST

0.89

best

0.88

 Best

0.88

 best

0.87

BEST

0.87

Activations Density 0.076%

expressions related to the concept of "best" or optimal choices

No Comments

No Known Activations