INDEX

Explanations

references to people and their accomplishments or backgrounds, especially in historical contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

 varandra

-0.86

 stället

-0.75

 igång

-0.70

 blest

-0.66

IntoConstraints

-0.65

 dourado

-0.63

 vägen

-0.63

 berätt

-0.61

 skydd

-0.61

 autorytatywna

-0.60

POSITIVE LOGITS

och

0.62

 för

0.59

upp

0.55

För

0.47

 För

0.47

ä

0.44

 tombé

0.43

är

0.42

av

0.42

 montr

0.42

Activations Density 0.084%

references to people and their accomplishments or backgrounds, especially in historical contexts

No Comments

No Known Activations