INDEX

Explanations

phrases indicating advice or recommendations for actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

OGND

-0.52

 Diretto

-0.49

 PeEnEo

-0.49

yym

-0.48

 jsonify

-0.47

agramm

-0.46

]++;

-0.46

遽

-0.46

Vidite

-0.46

Hj

-0.45

POSITIVE LOGITS

 remember

1.10

 consider

1.04

 beware

0.90

remember

0.86

 Consider

0.86

 Remember

0.85

 familiarize

0.82

 check

0.81

Remember

0.81

 avoid

0.80

Activations Density 0.282%

phrases indicating advice or recommendations for actions

No Comments

No Known Activations