INDEX

Explanations

sections of text that contain legal terminology or case references

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

featureID

-0.79

 mapStateToProps

-0.66

 comuniques

-0.65

writeFieldEnd

-0.65

 CURIAM

-0.65

ніципалі

-0.63

 Normdatei

-0.61

getDoctrine

-0.60

 MaterialApp

-0.60

RemoveField

-0.60

POSITIVE LOGITS

 chrétien

0.62

AndroidJUnit

0.57

ecake

0.54

 cherchez

0.52

 lastly

0.51

 glassy

0.50

 farmacé

0.50

0.49

 coupable

0.49

 Note

0.48

Activations Density 0.005%

sections of text that contain legal terminology or case references

No Comments

No Known Activations