© 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. 126051
    Prev
    Next
    INDEX
    Explanations

    charcoal/coal

    np_max-act · gemini-2.0-flash

    The neuron is tuned to detect mentions of activated charcoal (or “charcoal” in detox/adsorption contexts).

    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
     vive
    -0.07
     univers
    -0.06
    Parsing
    -0.06
     lays
    -0.06
     Ventura
    -0.06
    ')}}</
    -0.06
    _pres
    -0.06
    -dollar
    -0.06
     phân
    -0.06
     EDT
    -0.05
    POSITIVE LOGITS
     charcoal
    0.10
    mploy
    0.07
    朱
    0.07
    ublished
    0.07
    ジオ
    0.07
    อกจาก
    0.07
    ookie
    0.07
     cray
    0.07
    gil
    0.06
     paylaş
    0.06
    Activations Density 0.001%

    No Known Activations