© 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. 90733
    Prev
    Next
    INDEX
    Explanations

    true

    np_max-act · gemini-2.0-flash

    This neuron detects tokens that signal a film is based on real events—words like “true,” “life,” “based,” and “docu.”

    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
    cri
    -0.06
     Encyclopedia
    -0.06
     Garlic
    -0.06
    uts
    -0.06
    88
    -0.06
    建设
    -0.06
     Poe
    -0.06
    urd
    -0.06
    극
    -0.06
     greeting
    -0.06
    POSITIVE LOGITS
     cargo
    0.07
     рекоменда
    0.07
    Amt
    0.06
    ^.
    0.06
     bahwa
    0.06
    .header
    0.06
    PIC
    0.06
    أك
    0.06
     miscon
    0.06
     groupId
    0.06
    Activations Density 0.024%

    No Known Activations