INDEX

Explanations

comparative phrases indicating preference or distinction

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

"than."

np_acts-logits-general · gemini-2.0-flash

rather than

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 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

 Infórmanos

-0.66

 purpoſe

-0.62

༘

-0.58

 فريبيس

-0.56

 raiſ

-0.55

 ſta

-0.54

itanie

-0.54

,‎

-0.54

 houſe

-0.54

ſelf

-0.54

POSITIVE LOGITS

而不是

0.58

而非

0.55

 rather

0.54

 Rather

0.51

Rather

0.50

 zamiast

0.44

あくまで

0.44

rather

0.44

 plutôt

0.44

 вместо

0.41

Activations Density 0.298%

comparative phrases indicating preference or distinction

"than."

rather than

No Comments

No Known Activations