© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 41703
    Prev
    Next
    INDEX
    Explanations

    boards

    np_max-act · gemini-2.0-flash

    The neuron primarily activates on plural nouns (i.e., words marked with a plural ending).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_11/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.11.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Mansion
    -0.06
    -metal
    -0.06
     economics
    -0.06
    \R
    -0.06
     reactors
    -0.06
     secretive
    -0.06
     根
    -0.06
     inhibitors
    -0.06
    hips
    -0.06
    �
    -0.06
    POSITIVE LOGITS
    ->{_
    0.07
    >Total
    0.06
    stats
    0.06
    ่าเป
    0.06
    .onclick
    0.06
     tercih
    0.06
     půj
    0.06
     tales
    0.06
     akce
    0.06
    ائية
    0.05
    Activations Density 0.129%

    No Known Activations