INDEX

Explanations

terms related to the condition and quality of products

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

new/full/version following version/resolution/condition/grade

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

tanleria

-0.56

 advisors

-0.54

 fubject

-0.52

 fhew

-0.50

recated

-0.50

doc

-0.50

 Chwiliwch

-0.49

 myſelf

-0.49

 ſta

-0.48

sena

-0.48

POSITIVE LOGITS

 AssemblyTitle

0.47

KURZBESCHREIBUNG

0.46

 Qualität

0.41

OOTDTY

0.39

GenerationType

0.38

ExtendWith

0.36

ırlı

0.36

AutoScaleMode

0.35

 tarto

0.35

stdc

0.34

Activations Density 0.001%

terms related to the condition and quality of products

new/full/version following version/resolution/condition/grade

No Comments

No Known Activations