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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 41231
    Prev
    Next
    INDEX
    Explanations

    danger/harm

    np_max-act · gemini-2.0-flash

    This neuron detects words that convey danger or illegality (e.g., mentions of harm, being illegal, risk).

    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
     consequences
    -0.07
     caus
    -0.07
    callable
    -0.07
    ese
    -0.07
    escription
    -0.06
     Spanish
    -0.06
     renown
    -0.06
     FPS
    -0.06
    _enqueue
    -0.06
     name
    -0.06
    POSITIVE LOGITS
     litter
    0.07
     dém
    0.06
     الخ
    0.06
    <Entity
    0.06
     boob
    0.06
    CONTROL
    0.06
     ба
    0.06
     이동
    0.06
    链
    0.06
    .DisplayMember
    0.06
    Activations Density 0.028%

    No Known Activations