INDEX

Explanations

references to time and duration in various contexts

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

preceding "before"

np_acts-logits-general · gemini-2.0-flash

events before words like "before"

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

/*

-0.80

migrationBuilder

-0.66

WithIOException

-0.60

んぼ

-0.55

-0.50

IntoConstraints

-0.46

TintMode

-0.46

 chi̍t

-0.46

ItemBackground

-0.45

üstet

-0.45

POSITIVE LOGITS

 before

3.64

before

3.19

Before

2.95

 Before

2.94

 sebelum

2.85

 antes

2.80

 BEFORE

2.66

BEFORE

2.54

 voordat

2.54

 bevor

2.40

Activations Density 1.015%

references to time and duration in various contexts

preceding "before"

events before words like "before"

No Comments

No Known Activations