INDEX

Explanations

references to deceit and misinformation

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

Indicates something is not genuine or real

np_acts-logits-general · gemini-2.0-flash

claims that are false

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 24-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_24/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.24.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

wapV

-0.42

 Pill

-0.37

袱

-0.36

eneuve

-0.36

 TextAlign

-0.35

windowFixed

-0.35

StreetMap

-0.34

 vård

-0.34

 собі

-0.33

OGND

-0.33

POSITIVE LOGITS

 fake

0.91

 faked

0.87

fake

0.84

 phony

0.80

 faking

0.79

Fake

0.75

 fakes

0.74

 Fake

0.73

FAKE

0.71

 artificial

0.69

Activations Density 0.478%

references to deceit and misinformation

Indicates something is not genuine or real

claims that are false

No Comments

No Known Activations