INDEX

Explanations

references to academic authors and their affiliations or contributions in research papers

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

dates and numbers

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

 itſelf

-0.57

 Cæsar

-0.56

ſelf

-0.54

 fubject

-0.53

Etimología

-0.52

 Reſ

-0.52

性和

-0.52

 myſelf

-0.51

 ſche

-0.50

 pleaſure

-0.50

POSITIVE LOGITS

autoreleasepool

0.42

 nyelven

0.40

’.

0.39

'.

0.38

Datuak

0.38

 infine

0.37

 још

0.36

itattu

0.36

 llegaron

0.36

Activations Density 0.239%

references to academic authors and their affiliations or contributions in research papers

dates and numbers

No Comments

No Known Activations