INDEX

Explanations

phrases indicating error or wrongdoing

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_9/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.9.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

 vectorielles

-1.04

 tramonto

-0.98

 verksamhet

-0.92

AnchorStyles

-0.91

 attirer

-0.90

 énergé

-0.88

hadiran

-0.87

 vettore

-0.85

 mattino

-0.85

 quelcon

-0.85

POSITIVE LOGITS

 wrong

1.48

 WRONG

1.42

 Wrong

1.40

WRONG

1.33

wrong

1.30

Wrong

1.26

 wrongs

0.93

 conci

0.82

 incorrect

0.80

 错误

0.76

Activations Density 0.068%

phrases indicating error or wrongdoing

No Comments

No Known Activations