INDEX

Explanations

phrases indicating personal experiences and opinions about recognition

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

Follows "never," "ve" or "have" and precedes "heard"

np_acts-logits-general · gemini-2.0-flash

heard or not heard

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

__":

-0.55

SharedDtor

-0.52

رشف

-0.50

▾

-0.45

tetés

-0.45

strijd

-0.45

 Respectfully

-0.44

yntaxException

-0.44

 Verfügung

-0.43

XDECREF

-0.43

POSITIVE LOGITS

 heard

2.56

 hear

2.33

 hears

2.13

 hearing

2.11

heard

2.02

Heard

1.90

 Heard

1.83

 HEAR

1.75

 Hear

1.72

hearing

1.71

Activations Density 0.246%

phrases indicating personal experiences and opinions about recognition

Follows "never," "ve" or "have" and precedes "heard"

heard or not heard

No Comments

No Known Activations