INDEX

Explanations

the word "before" and its variations, indicating a focus on temporal relationships or sequences

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

"before"

np_acts-logits-general · gemini-2.0-flash

temporal sequence before

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_9/width_16k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.9.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

 partielle

-0.47

 touristique

-0.46

 koszulka

-0.46

 culturelles

-0.46

 récents

-0.45

 Empfindung

-0.44

 miniaturka

-0.43

 eiland

-0.43

 Tourisme

-0.43

 flotante

-0.43

POSITIVE LOGITS

 before

2.00

before

1.89

 Before

1.71

Before

1.70

 BEFORE

1.62

BEFORE

1.57

 sebelum

1.53

 antes

1.46

 før

1.34

 înainte

1.30

Activations Density 0.130%

the word "before" and its variations, indicating a focus on temporal relationships or sequences

"before"

temporal sequence before

No Comments

No Known Activations