INDEX

Explanations

references to quantitative comparisons or statistics

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

timestamps and dates

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

Personendaten

-0.60

OCCURRED

-0.49

bcryptjs

-0.48

 torchvision

-0.46

 Arjuna

-0.45

 Dhoni

-0.45

 saites

-0.43

LinkId

-0.43

 newOwner

-0.43

 Picchu

-0.42

POSITIVE LOGITS

 cremallera

0.46

ValueStyle

0.42

Reply

0.37

MLLoader

0.36

mallera

0.36

 kemeja

0.36

 vänner

0.35

 متعلقه

0.34

VYMaps

0.33

 Gespräche

0.33

Activations Density 0.005%

references to quantitative comparisons or statistics

timestamps and dates

No Comments

No Known Activations