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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 58805
    Prev
    Next
    INDEX
    Explanations

    head

    np_max-act · gemini-2.0-flash

    The neuron activates on the token “head” (especially capitalized “Head”), identifying occurrences of that word in headings or technical contexts.

    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
    Uri
    -0.07
    iltr
    -0.07
    _uri
    -0.07
    INTR
    -0.06
     contraseña
    -0.06
    �
    -0.06
    _xor
    -0.06
    oric
    -0.06
    atır
    -0.06
     dafür
    -0.06
    POSITIVE LOGITS
     head
    0.16
    head
    0.15
     Head
    0.14
    HEAD
    0.13
    Head
    0.12
     heads
    0.12
     HEAD
    0.12
    -head
    0.11
     Heads
    0.11
     heading
    0.10
    Activations Density 0.033%

    No Known Activations