INDEX

Explanations

phrases related to deception and misleading actions

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

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

期刊论文

-0.49

PostExecute

-0.44

 jaula

-0.43

 élev

-0.42

 suyo

-0.41

 <=",

-0.40

RectangleBorder

-0.40

 larmes

-0.40

 échanges

-0.40

 Geschle

-0.39

POSITIVE LOGITS

 viewers

0.56

 voters

0.54

 listeners

0.53

 readers

0.52

 audiences

0.51

 investors

0.48

informed

0.47

 consumers

0.46

us

0.46

 Ghanaians

0.44

Activations Density 0.460%

phrases related to deception and misleading actions

No Comments

No Known Activations