INDEX

Explanations

code snippets, especially in Java and Swift

oai_token-act-pair · gemini-2.0-flash

extends

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_0/width_16k/average_l0_76

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

 myſelf

-1.79

 itſelf

-1.78

Efq

-1.69

 doubtnut

-1.57

 pleaſure

-1.55

 Jefus

-1.53

 faſt

-1.53

 Majefty

-1.49

 houſe

-1.48

 Monfieur

-1.48

POSITIVE LOGITS

<eos>

1.76

↵↵

1.69

↵↵↵

1.42

↵

1.33

↵↵↵↵

1.23

↵↵↵↵↵

1.05

...

1.03

↵↵↵↵↵↵

0.97

↵↵↵↵↵↵↵

0.93

</blockquote>

0.91

Activations Density 1.909%

code snippets, especially in Java and Swift

extends

No Comments

No Known Activations