INDEX

Explanations

references to educational institutions and related historical context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_18/width_65k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.18.hook_mlp_out

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 старости

-0.41

retweet

-0.41

astronomie

-0.41

chability

-0.39

 Botox

-0.38

साय

-0.37

writel

-0.37

setVerticalGroup

-0.37

gh

-0.36

 Sire

-0.36

POSITIVE LOGITS

 named

2.05

 Named

1.95

named

1.92

Named

1.87

 honour

1.51

 naming

1.45

 honor

1.44

 names

1.35

 name

1.27

 nommée

1.27

Activations Density 0.219%

references to educational institutions and related historical context

No Comments

No Known Activations