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. 86274
    Prev
    Next
    INDEX
    Explanations

    software errors

    np_max-act · gemini-2.0-flash

    This neuron detects numeric version or parameter values—especially decimal numbers (e.g. “1.4.9”, “0.6020…”, library or tool version numbers).

    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
    osals
    -0.07
    Aura
    -0.07
     Medi
    -0.06
    .centerX
    -0.06
     فاصله
    -0.06
     Το
    -0.06
    광
    -0.06
    suz
    -0.06
    _tag
    -0.06
    Low
    -0.06
    POSITIVE LOGITS
    _SEGMENT
    0.07
    ↵
    0.06
     ايران
    0.06
     Murray
    0.06
     tả
    0.06
    _HOOK
    0.06
     nec
    0.06
     BT
    0.06
     processors
    0.06
     chinese
    0.06
    Activations Density 0.006%

    No Known Activations