INDEX

Explanations

place names and names of people and organizations involved in politics or crime

oai_token-act-pair · gemini-2.0-flash

political/legal events

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_22/width_16k/average_l0_15

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

-0.42

seu

-0.38

Prev

-0.37

 förm

-0.37

 previous

-0.36

cat

-0.36

<eos>

-0.35

 cabe

-0.35

ce

-0.34

POSITIVE LOGITS

Efq

0.95

 myſelf

0.91

 Monfieur

0.91

 itſelf

0.90

ſelf

0.88

 themſelves

0.87

 Houſe

0.86

 ſind

0.85

 correctes

0.84

 houſe

0.84

Activations Density 4.046%

place names and names of people and organizations involved in politics or crime

political/legal events

No Comments

No Known Activations