INDEX

Explanations

interactions and inquiries about various characters or subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 17-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_17/width_65k/average_l0_125

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.17.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

 ivelany

-0.68

hObject

-0.65

XmlEnum

-0.59

 Wikimédia

-0.57

ίκη

-0.55

SPIRE

-0.55

mtrl

-0.52

rittura

-0.52

TextBoxColumn

-0.52

 Useful

-0.52

POSITIVE LOGITS

 questions

0.97

 question

0.84

questions

0.83

 whether

0.81

question

0.75

 perguntas

0.73

 pregunta

0.73

 Questions

0.70

Fragen

0.69

 preguntas

0.68

Activations Density 0.036%

interactions and inquiries about various characters or subjects

No Comments

No Known Activations