© 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. 125162
    Prev
    Next
    INDEX
    Explanations

    love, sex, intimacy

    np_max-act · gemini-2.0-flash

    sensual descriptions of intimate physical interactions between characters.

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

    The neuron primarily detects and fires on placeholder character name tokens (e.g. NAME_1, NAME_2) in the text.

    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
     duo
    -0.06
    opped
    -0.06
     lt
    -0.06
     EFI
    -0.06
     decad
    -0.06
     Á
    -0.06
     curb
    -0.06
    icing
    -0.05
    .worker
    -0.05
     Με
    -0.05
    POSITIVE LOGITS
     spindle
    0.07
     sculpture
    0.07
     ans
    0.07
     confused
    0.06
     touched
    0.06
     disease
    0.06
     Fransa
    0.06
     restau
    0.06
     행동
    0.06
     testing
    0.06
    Activations Density 0.016%

    No Known Activations