INDEX

Explanations

terms related to online privacy and security tools, particularly VPNs and proxies

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

anonymous use of 'tor'

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

 ✭✭

-0.51

Dea

-0.43

 plaintiff

-0.39

TargetException

-0.39

artifactId

-0.38

DockStyle

-0.36

auté

-0.36

getDoctrine

-0.35

被告

-0.35

IBOutlet

-0.35

POSITIVE LOGITS

 nakalista

0.50

 Gegenteil

0.45

 busto

0.44

privacy

0.40

versations

0.38

 seyn

0.38

 Konsequ

0.38

ItemBackground

0.38

EFIAPI

0.37

học

0.37

Activations Density 0.035%

terms related to online privacy and security tools, particularly VPNs and proxies

anonymous use of 'tor'

No Comments

No Known Activations