INDEX

Explanations

pronouns and personal references

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

Precedes "want" or a form of "want"

np_acts-logits-general · gemini-2.0-flash

expressing desire or need

np_acts-logits-general · gemini-2.5-flash-lite

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 للاسماء

-0.47

ensement

-0.46

Réponse

-0.43

uté

-0.41

 localObject

-0.39

foque

-0.38

tasking

-0.38

ometric

-0.37

jspx

-0.37

wej

-0.37

POSITIVE LOGITS

 want

3.30

want

2.90

 wants

2.85

 Want

2.76

 wanting

2.67

Want

2.57

 wanted

2.54

 WANT

2.41

wants

2.37

 Wants

2.29

Activations Density 0.741%

pronouns and personal references

Precedes "want" or a form of "want"

expressing desire or need

No Comments

No Known Activations