INDEX

Explanations

phrases that indicate change or qualification in statements

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

increasing resentment, excessive force, limited success

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 sinar

-0.43

 liné

-0.41

 grève

-0.39

 calorías

-0.37

RTLI

-0.36

ğraf

-0.36

ändigkeit

-0.36

 Normdatei

-0.35

 circulaire

-0.35

 nadie

-0.35

POSITIVE LOGITS

rungsseite

0.67

AlterField

0.57

RenderAtEndOf

0.54

SourceChecksum

0.53

ConstraintMaker

0.52

XtraReports

0.51

曖昧さ回避

0.49

 незавершена

0.47

ptăm

0.47

AndView

0.46

Activations Density 0.247%

phrases that indicate change or qualification in statements

increasing resentment, excessive force, limited success

No Comments

No Known Activations