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. 86310
    Prev
    Next
    INDEX
    Explanations

    Python errors with line numbers

    np_max-act · gemini-2.0-flash

    The neuron flags numeric tokens—especially floating‐point numbers (e.g. line‐number values with decimals) in stack traces.

    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
    _sm
    -0.06
    intree
    -0.06
    ronics
    -0.06
     railway
    -0.06
     TMP
    -0.06
    uj
    -0.06
    -star
    -0.06
     Yük
    -0.06
    ені
    -0.06
     ด
    -0.06
    POSITIVE LOGITS
    institution
    0.07
    ↵↵↵↵↵↵
    0.07
     устройства
    0.07
    	UPROPERTY
    0.07
    Specific
    0.06
    永
    0.06
    errorMsg
    0.06
     благод
    0.06
    bucket
    0.06
    Acknowled
    0.06
    Activations Density 0.005%

    No Known Activations