© 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. 13443
    Prev
    Next
    INDEX
    Explanations

    convention

    np_max-act · gemini-2.0-flash

    The neuron activates on instances of the word “Conventionally” (i.e. mentions of conventional or prior‐art approaches).

    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
    ...'↵
    -0.07
    -blood
    -0.07
    RequestBody
    -0.07
     unterstüt
    -0.07
    income
    -0.06
    140
    -0.06
     aides
    -0.06
    ecektir
    -0.06
     Armed
    -0.06
    Smoke
    -0.06
    POSITIVE LOGITS
     convention
    0.14
     conventions
    0.12
     Convention
    0.11
    Convention
    0.10
     onUpdate
    0.07
     haven
    0.07
    NV
    0.07
     Env
    0.07
     heaven
    0.07
     honeymoon
    0.07
    Activations Density 0.006%

    No Known Activations