INDEX

Explanations

various conjunctions and quantifiers often found in lists or descriptions

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

list of acronyms or initialisms

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

 nonUne

-0.38

 dépenses

-0.37

 gouvernements

-0.35

 Taktlose

-0.34

job

-0.33

 étrangers

-0.33

 obligation

-0.32

 western

-0.32

 economica

-0.32

setColumns

-0.31

POSITIVE LOGITS

 مرئيه

0.64

AndEndTag

0.54

 محفوظة

0.53

TypedDataSet

0.52

 berjudul

0.50

ContentAsync

0.50

還有

0.49

Efq

0.48

 Majefty

0.48

 dilan

0.47

Activations Density 0.057%

various conjunctions and quantifiers often found in lists or descriptions

list of acronyms or initialisms

No Comments

No Known Activations