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. 54788
    Prev
    Next
    INDEX
    Explanations

    discrimination, hate

    np_max-act · gemini-2.0-flash

    content related to harmful behavior and hate speech.

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

    This neuron detects words related to harmful or disallowed content—e.g., hate, speech, discrimination, harmful.

    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
    िस
    -0.06
    riteln
    -0.06
    ToDelete
    -0.06
    okedex
    -0.06
     зм
    -0.06
     rng
    -0.06
    Paused
    -0.06
    .te
    -0.06
     여자
    -0.06
    _orient
    -0.06
    POSITIVE LOGITS
     democratic
    0.06
    wan
    0.06
     فعالیت
    0.06
     TOM
    0.06
    ΐ
    0.06
     بلغ
    0.06
     cos
    0.06
    abbit
    0.06
     masterpiece
    0.06
     BigInteger
    0.06
    Activations Density 0.016%

    No Known Activations