INDEX

Explanations

references to time, specifically different expressions of the past

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.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

 прошло

-0.39

teles

-0.39

的一次

-0.38

Пото

-0.37

 屋根

-0.36

embra

-0.35

もちゃ

-0.35

ToOne

-0.35

一样的

-0.35

POSITIVE LOGITS

few

1.13

 couple

0.93

decade

0.83

 dozen

0.72

two

0.71

few

0.69

 five

0.68

 three

0.68

 handful

0.68

 seven

0.67

Activations Density 0.156%

references to time, specifically different expressions of the past

No Comments

No Known Activations

references to time, specifically different expressions of the past

No Comments

No Known Activations