© 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. 30529
    Prev
    Next
    INDEX
    Explanations

    quality and performance

    np_max-act · gemini-2.0-flash

    The neuron consistently activates on terms referring to quality‐control and performance metrics (e.g. “quality,” “control,” “performance,” “precision,” “QC”) in analytical or laboratory 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
    urse
    -0.06
     inputs
    -0.06
    _CAM
    -0.06
    átor
    -0.06
     looming
    -0.06
     Curve
    -0.06
    番
    -0.06
    agogue
    -0.06
    陣
    -0.06
    pass
    -0.06
    POSITIVE LOGITS
     island
    0.07
    >");
↵
    0.07
    avigate
    0.06
    ',

↵
    0.06
     poate
    0.06
    _Model
    0.06
     Rey
    0.06
    ()");↵
    0.06
    ै?
    0.06
    *dx
    0.06
    Activations Density 0.019%

    No Known Activations