INDEX

Explanations

statements of self-assessment regarding expertise and knowledge in a particular subject

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

not an expert

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

Demografía

-0.70

 Wikimedijinoj

-0.66

 betweenstory

-0.61

 desmotivaciones

-0.57

BoxShadow

-0.56

 Photocase

-0.54

astify

-0.53

 NSCoder

-0.52

inWeight

-0.51

 marbre

-0.51

POSITIVE LOGITS

 EXPERT

0.46

 Experts

0.44

 Expert

0.43

Experts

0.41

 experts

0.40

 expert

0.39

Expert

0.39

expert

0.38

fag

0.37

 EXPER

0.36

Activations Density 0.024%

statements of self-assessment regarding expertise and knowledge in a particular subject

not an expert

No Comments

No Known Activations