© 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. 126613
    Prev
    Next
    INDEX
    Explanations

    ----------------

    np_max-act · gemini-2.0-flash

    The neuron activates on long runs of dashes (separator lines) used to mark boundaries (e.g. forward/divider lines) in email text.

    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
    ;o
    -0.08
    .xpath
    -0.07
     Comm
    -0.07
    cba
    -0.07
    /GL
    -0.06
    electron
    -0.06
    "c
    -0.06
    .man
    -0.06
    Bin
    -0.06
    řet
    -0.06
    POSITIVE LOGITS
    ////////////////////////////////
    0.06
    ----------------
    0.06
     accelerator
    0.06
    .ForeignKey
    0.06
    ForeignKey
    0.06
    irtschaft
    0.06
     tomar
    0.06
    piel
    0.06
     resolver
    0.06
    Bước
    0.06
    Activations Density 0.000%

    No Known Activations