© 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. 6565
    Prev
    Next
    INDEX
    Explanations

    on/off, active/inactive

    np_max-act · gemini-2.0-flash

    The neuron fires on terms that describe control actions or state changes—words like “activation,” “actuated,” “closed,” “sequencing,” “deactivated,” etc.

    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
    ibrator
    -0.08
    seven
    -0.07
     Laboratory
    -0.07
    addon
    -0.07
     самое
    -0.06
    (pass
    -0.06
     hakk
    -0.06
    ریب
    -0.06
     stencil
    -0.06
    耳
    -0.06
    POSITIVE LOGITS
    -files
    0.07
    	Runtime
    0.07
     сум
    0.07
    σιεύ
    0.06
     FILES
    0.06
     Фор
    0.06
     Lim
    0.06
    /">
    0.06
     zvol
    0.06
    باز
    0.06
    Activations Density 0.026%

    No Known Activations