© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 92500
    Prev
    Next
    INDEX
    Explanations

    im

    np_max-act · gemini-2.0-flash

    programming code and syntax elements.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    The neuron fires on ordinary natural‐language words (e.g. English or Latin content words) and stays off for code or symbols.

    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
    زارش
    -0.07
    oS
    -0.07
    들에게
    -0.07
    сит
    -0.07
    _help
    -0.06
    	s
    -0.06
    swer
    -0.06
    áhl
    -0.06
    商
    -0.06
    (turn
    -0.06
    POSITIVE LOGITS
    科学
    0.07
     Offer
    0.06
    .neo
    0.06
     Exercises
    0.06
    Interior
    0.06
     consequently
    0.06
     medios
    0.06
     konus
    0.06
     consciously
    0.06
     Accom
    0.06
    Activations Density 0.005%

    No Known Activations