INDEX

Explanations

phrases related to spending time with friends and family, especially "hanging out".

oai_token-act-pair · gemini-2.0-flash

hang

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_19/width_16k/average_l0_12

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.19.ln2.hook_normalized

Architecture

jumprelu_transcoder

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 proposal

-1.30

 Proposal

-1.16

proposal

-1.10

Proposal

-1.08

 propose

-1.07

 proposals

-1.03

 proposing

-1.01

 PROPOS

-0.96

 Proposals

-0.95

 proposes

-0.90

POSITIVE LOGITS

 hang

0.68

 Hang

0.65

 HANG

0.64

Hang

0.60

HANG

0.59

 around

0.57

 fooling

0.55

 hanging

0.54

 OnInit

0.53

 loadImage

0.52

Activations Density 0.387%

phrases related to spending time with friends and family, especially "hanging out".

hang

No Comments

No Known Activations