INDEX

Explanations

proper nouns and significant numeric values related to events

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

-0.46

 Infórmanos

-0.45

findpost

-0.44

CELLANEOUS

-0.44

 désolés

-0.42

tagHelperRunner

-0.41

*/].

-0.36

Jas

-0.36

 cherchés

-0.35

ellido

-0.35

POSITIVE LOGITS

 AssemblyCulture

0.50

 gantung

0.47

 vaksin

0.47

ActionMode

0.47

 palsu

0.46

 AssemblyVersion

0.45

 člán

0.44

 sempurna

0.44

 garansi

0.44

 estatua

0.43

Activations Density 0.026%

proper nouns and significant numeric values related to events

No Comments

No Known Activations