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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 11-RESID-POST-AA
    6. 130564
    Prev
    Next
    INDEX
    Explanations

    violence

    np_max-act · gemini-2.0-flash

    This neuron activates on numeric tokens—especially numbers with decimal points (floating-point values).

    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
     oblast
    -0.07
     �
    -0.06
    First
    -0.06
     ----------------------------------------------------------------------↵
    -0.06
     Đức
    -0.06
     Zuk
    -0.06
     kiss
    -0.06
    ↵            ↵
    -0.06
     zu
    -0.06
    /the
    -0.06
    POSITIVE LOGITS
    _PEER
    0.07
     authToken
    0.07
    _COMMON
    0.07
    	work
    0.07
    υγ
    0.07
     intros
    0.06
    cstdlib
    0.06
    TP
    0.06
    ☴
    0.06
    anceled
    0.06
    Activations Density 0.038%

    No Known Activations