INDEX

Explanations

details about special edition media releases

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

Following "as," "authors," or "reveals."

np_acts-logits-general · gemini-2.0-flash

behind

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

guém

-0.53

idän

-0.51

rungsseite

-0.51

+#+#

-0.49

 répertoire

-0.48

ValueGeneration

-0.47

 výraz

-0.46

 createState

-0.46

pakatan

-0.46

Gambas

-0.45

POSITIVE LOGITS

 behind

1.18

behind

1.08

 Behind

1.02

Behind

0.98

 backstage

0.98

 bonus

0.96

 BEHIND

0.95

 detrás

0.85

bonus

0.80

 Bonus

0.80

Activations Density 0.235%

details about special edition media releases

Following "as," "authors," or "reveals."

behind

No Comments

No Known Activations