INDEX

Explanations

terms related to originality and authentic works

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

"original" before a noun

np_acts-logits-general · gemini-2.0-flash

original followed by noun

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

ValueStyle

-0.63

 дописавши

-0.63

 vPvB

-0.60

bbero

-0.58

 enkelte

-0.58

THINGS

-0.57

 ziua

-0.57

 altrimenti

-0.56

 âgées

-0.55

 officiellement

-0.55

POSITIVE LOGITS

 intention

0.78

 intended

0.74

 Intended

0.73

 intent

0.73

 Original

0.72

 source

0.72

splan

0.72

 заду

0.72

 ORIGINAL

0.71

 orig

0.69

Activations Density 0.065%

terms related to originality and authentic works

"original" before a noun

original followed by noun

No Comments

No Known Activations