© 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. 55537
    Prev
    Next
    INDEX
    Explanations

    long sentences

    np_max-act · gemini-2.0-flash

    The neuron responds to relatively long content tokens—i.e. it activates on multi‐character (longer) words or word fragments rather than short/common tokens.

    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
    Tax
    -0.07
    ++;
↵
↵
    -0.06
     Five
    -0.06
    [y
    -0.06
    jist
    -0.06
    _ad
    -0.06
    52
    -0.06
     Cơ
    -0.06
    -ves
    -0.06
    orny
    -0.06
    POSITIVE LOGITS
    ','=','
    0.06
    appers
    0.06
     empower
    0.06
    bootstrap
    0.06
     Specify
    0.06
    لال
    0.06
    (number
    0.06
    monitor
    0.06
    Compilation
    0.06
    :],
    0.06
    Activations Density 0.026%

    No Known Activations