INDEX

Explanations

references to violent incidents and casualties

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 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

errHandler

-0.64

PyExc

-0.54

Hrsg

-0.50

deepcopy

-0.47

 للمعارف

-0.46

ELA

-0.45

WindowEvent

-0.45

 évi

-0.43

ByPrimaryKey

-0.42

hassee

-0.41

POSITIVE LOGITS

 injured

0.85

 Injured

0.81

 wounded

0.75

INCLUDING

0.70

TagMode

0.70

 JUGA

0.70

 heridos

0.69

 injure

0.69

injured

0.69

 injuries

0.68

Activations Density 0.163%

references to violent incidents and casualties

No Comments

No Known Activations