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. 16177
    Prev
    Next
    INDEX
    Explanations

    Scientific abbreviations

    np_max-act · gemini-2.0-flash

    The neuron activates on short, all-caps scientific initialisms—especially biochemical compound or lipid acronyms like FA, TG, VLDL, HA, 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
     ô
    -0.07
    illin
    -0.07
    cuda
    -0.07
     meisje
    -0.07
     servlet
    -0.07
    Picture
    -0.07
    雅
    -0.06
     experiments
    -0.06
    Index
    -0.06
    otypes
    -0.06
    POSITIVE LOGITS
    issional
    0.06
    ambah
    0.06
    -describedby
    0.06
     než
    0.05
    گي
    0.05
    асти
    0.05
    από
    0.05
     Loop
    0.05
    vable
    0.05
     columnIndex
    0.05
    Activations Density 0.044%

    No Known Activations