INDEX

Explanations

punctuation marks and sentence boundaries

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

possibility and elaboration

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

 CreateTagHelper

-0.48

 fieldNum

-0.46

djangoproject

-0.44

expandindo

-0.44

 estekak

-0.44

 docteur

-0.42

 pleaſure

-0.42

 commandement

-0.42

❉

-0.41

 basically

-0.40

POSITIVE LOGITS

 might

0.76

might

0.66

anskje

0.64

posedge

0.63

may

0.61

 måske

0.61

 kanskje

0.59

 évent

0.58

 möglicherweise

0.58

 Might

0.57

Activations Density 0.137%

punctuation marks and sentence boundaries

possibility and elaboration

No Comments

No Known Activations