INDEX

Explanations

repeated sequences or patterns within text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

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

 BoxDecoration

-0.84

httphttps

-0.74

IndentedString

-0.73

 initComponents

-0.72

adaptiveStyles

-0.69

Personendaten

-0.68

 ExecuteAsync

-0.68

parsedMessage

-0.68

 autorytatywna

-0.66

+#+#

-0.66

POSITIVE LOGITS

0.47

 hence

0.43

 therefore

0.41

 capaces

0.41

 perhaps

0.40

 cuya

0.39

 whose

0.39

Like

0.39

 cuyo

0.38

 thus

0.38

Activations Density 0.103%

repeated sequences or patterns within text

No Comments

No Known Activations