INDEX

Explanations

instances of deception and misinformation

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

Dishonesty, deception, or misrepresentation

np_acts-logits-general · gemini-2.0-flash

lying, falsehoods, misrepresentation

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 transfieras

-0.43

 Geschmack

-0.40

 taste

-0.38

úgó

-0.37

 jäsen

-0.36

ClientRect

-0.34

 Cristal

-0.34

horabuena

-0.34

PathParam

-0.33

Inactive

-0.33

POSITIVE LOGITS

InputTagHelper

0.71

aarrggbb

0.59

 AspNetCore

0.56

 linkovi

0.56

 lied

0.55

 concealing

0.52

principalColumn

0.52

 estekak

0.50

 AssemblyCulture

0.49

tanleria

0.49

Activations Density 0.459%

instances of deception and misinformation

Dishonesty, deception, or misrepresentation

lying, falsehoods, misrepresentation

No Comments

No Known Activations