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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 9499
    Prev
    Next
    INDEX
    Explanations

    adjustable systems

    np_max-act · gemini-2.0-flash

    The neuron activates on verbs and noun phrases that describe adjusting or controlling mechanical output (e.g., throttle, regulate, change, control, adjust).

    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
    Once
    -0.07
     fades
    -0.06
    ože
    -0.06
     stout
    -0.06
    counter
    -0.06
     blows
    -0.06
    [curr
    -0.06
    іту
    -0.06
    орони
    -0.06
    shutdown
    -0.06
    POSITIVE LOGITS
    bere
    0.07
     Οι
    0.06
     살아
    0.06
    ISS
    0.06
    pheric
    0.06
     kalp
    0.06
    IGINAL
    0.06
    �
    0.06
     زنی
    0.06
    .DateFormat
    0.06
    Activations Density 0.024%

    No Known Activations