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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 100521
    Prev
    Next
    INDEX
    Explanations

    how to

    np_max-act · gemini-2.0-flash

    mentions or requests for instructions on making explosives or other harmful devices.

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0

    requests or instructions related to making explosives or bombs, especially “how to make” style queries and recipes.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    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
    .YES
    -0.07
     detects
    -0.06
    mph
    -0.06
    _Port
    -0.06
    ห
    -0.06
    stitial
    -0.06
    _Mod
    -0.06
    _top
    -0.06
     المنت
    -0.06
     pride
    -0.06
    POSITIVE LOGITS
     â
    0.07
    utors
    0.06
    Throughout
    0.06
    	signal
    0.06
    _DECREF
    0.06
    BlockSize
    0.06
    EventType
    0.06
    ////////////////////////////////////////////////////////////////////////////////////////////////
    0.06
    alytics
    0.06
    (Optional
    0.06
    Activations Density 0.035%

    No Known Activations