INDEX

Explanations

several patterns, but the most relevant seems to be numbers and related words like "both", "first", "second", "one", "two"

oai_token-act-pair · gemini-2.0-flash

both

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_18/width_16k/average_l0_13

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

fjspx

-0.80

 AssemblyCulture

-0.76

 resourceCulture

-0.75

 prochaines

-0.69

ksesta

-0.68

 âmes

-0.66

EndGlobalSection

-0.64

adpleegd

-0.64

 snippetHide

-0.63

 vœux

-0.61

POSITIVE LOGITS

 counterparts

0.43

 situations

0.43

 candidates

0.42

 currencies

0.42

 versions

0.38

 grat

0.37

 friends

0.37

 scenarios

0.37

 alternatives

0.36

 issues

0.36

Activations Density 4.218%

several patterns, but the most relevant seems to be numbers and related words like "both", "first", "second", "one", "two"

both

No Comments

No Known Activations