INDEX

Explanations

references to deception or false appearances

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

deception, not genuine, illusory

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

 klubben

-0.43

autorest

-0.43

 sáb

-0.39

quelles

-0.37

 preferencia

-0.37

 vuestros

-0.35

+#+#

-0.34

 Kebijakan

-0.34

OneToMany

-0.34

뀜

-0.34

POSITIVE LOGITS

 fake

0.83

 Fake

0.77

fake

0.74

 sham

0.73

 bogus

0.72

 falsa

0.69

 phony

0.68

 fakes

0.66

 hoax

0.66

 false

0.65

Activations Density 0.039%

references to deception or false appearances

deception, not genuine, illusory

No Comments

No Known Activations