INDEX

Explanations

conditional and suggestive phrases indicating potential actions or outcomes

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

uncertainty or possibility

np_acts-logits-general · gemini-2.0-flash

modal verbs indicating possibility

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

 fieldNum

-0.50

@"/

-0.49

ReusableCell

-0.47

ArgumentParser

-0.45

 CreateTagHelper

-0.45

FormTagHelper

-0.45

ficulties

-0.41

قایناقلار

-0.41

transQ

-0.41

 colic

-0.40

POSITIVE LOGITS

 AssemblyVersion

0.53

 ویکی‌پدی

0.47

hopefully

0.44

 jöv

0.43

 Hopefully

0.43

 säkert

0.43

 hopefully

0.43

Portail

0.42

دانشنامهٔ

0.42

oa̍t

0.42

Activations Density 0.827%

conditional and suggestive phrases indicating potential actions or outcomes

uncertainty or possibility

modal verbs indicating possibility

No Comments

No Known Activations