© 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
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 73721
    Prev
    Next
    INDEX
    Explanations

    technology usage

    np_max-act · gemini-2.0-flash

    The neuron selectively activates on parts of floating-point numbers—i.e. the decimal tokens that represent probability or activation values.

    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
     chicks
    -0.08
    ाण
    -0.07
     trustee
    -0.07
     picnic
    -0.07
    aks
    -0.07
    Bush
    -0.07
     جهان
    -0.07
    omial
    -0.07
     століття
    -0.06
     gpu
    -0.06
    POSITIVE LOGITS
     <--
    0.06
    .setString
    0.06
     Yıl
    0.06
    .Translate
    0.06
     haystack
    0.06
    converter
    0.06
     مات
    0.06
    .rad
    0.06
     учеб
    0.06
    pgsql
    0.06
    Activations Density 0.076%

    No Known Activations