© 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. 125953
    Prev
    Next
    INDEX
    Explanations

    reduce

    np_max-act · gemini-2.0-flash

    The neuron detects words that signal degradation, reduction, or failure—i.e. terms indicating negative impacts or performance declines.

    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
    isd
    -0.06
    Included
    -0.06
    .energy
    -0.06
     perpetrators
    -0.06
     houses
    -0.06
     agosto
    -0.06
     Ontario
    -0.06
     yaptığı
    -0.06
    .Ed
    -0.06
     Πλη
    -0.06
    POSITIVE LOGITS
    :string
    0.07
    ган
    0.07
     scrim
    0.07
     ----------↵
    0.06
    [sub
    0.06
    авлива
    0.06
     extingu
    0.06
     lexer
    0.06
    ([])↵
    0.06
     لب
    0.06
    Activations Density 0.055%

    No Known Activations