INDEX

Explanations

expressions of potential legal violations or misconduct

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

Appears before a number

np_acts-logits-general · gemini-2.0-flash

RONALD MIXON

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

ʺ

-0.92

Tikang

-0.87

 بيها

-0.84

 متعلقه

-0.79

 ujednoznacz

-0.79

 تضيفلها

-0.72

AnchorStyles

-0.72

Kw

-0.68

heets

-0.67

styleType

-0.65

POSITIVE LOGITS

↵↵↵↵

1.73

↵↵↵↵↵

1.55

↵↵↵↵↵↵

1.52

↵↵↵

1.52

↵↵↵↵↵↵↵

1.38

↵↵↵↵↵↵↵↵

1.33

↵↵↵↵↵↵↵↵↵

1.28

↵↵↵↵↵↵↵↵↵↵

1.23

↵↵↵↵↵↵↵↵↵↵↵

1.22

↵↵↵↵↵↵↵↵↵↵↵↵

1.10

Activations Density 0.056%

expressions of potential legal violations or misconduct

Appears before a number

RONALD MIXON

No Comments

No Known Activations