INDEX

Explanations

phrases and terms related to negative traits or characteristics of people or groups

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

expandindo

-0.45

 muualla

-0.45

 вікісторінку

-0.44

❹

-0.41

OGND

-0.40

ngdoc

-0.40

xase

-0.39

ArrowToggle

-0.39

tagHelperRunner

-0.39

VolleyError

-0.38

POSITIVE LOGITS

 irritable

0.54

 fickle

0.51

 unruly

0.50

 fussy

0.50

 stubborn

0.50

 behavior

0.49

 humanidade

0.47

behavior

0.47

 obstinate

0.46

 troublesome

0.46

Activations Density 0.121%

phrases and terms related to negative traits or characteristics of people or groups

No Comments

No Known Activations