Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 7-RESID-POST-AA
    6. 55221
    Prev
    Next
    INDEX
    Explanations

    acronyms

    np_max-act · gemini-2.0-flash

    This neuron responds to uppercase acronyms, initialisms, or all-caps letter sequences (capitalized token fragments).

    oai_token-act-pair · gpt-5-miniTriggered by @yooniel31
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_7/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.7.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    /Object
    -0.07
    .Icon
    -0.07
    398
    -0.06
    Forecast
    -0.06
    ειτουργ
    -0.06
     Pap
    -0.06
    antity
    -0.06
    448
    -0.06
    _MULT
    -0.06
     Pepper
    -0.06
    POSITIVE LOGITS
     तह
    0.07
    EF
    0.07
    θή
    0.07
    								  
    0.06
     kendisine
    0.06
     seul
    0.06
    ={{
    0.06
    usuarios
    0.06
    议
    0.06
     üzerine
    0.06
    Activations Density 0.206%

    No Known Activations