INDEX

Explanations

expressions of commendation or praise

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

etra

-0.58

into

-0.55

 Levin

-0.53

Slo

-0.52

ne

-0.52

nage

-0.51

зму

-0.51

__))

-0.50

 costi

-0.50

POSITIVE LOGITS

 praise

1.87

 praises

1.74

 praising

1.72

 praised

1.71

praise

1.70

 applaud

1.54

 Praise

1.53

 commend

1.48

 commendation

1.43

Praise

1.39

Activations Density 0.163%

expressions of commendation or praise

No Comments

No Known Activations