Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 85860
    Prev
    Next
    INDEX
    Explanations

    Grammar and language usage

    np_max-act · gemini-2.0-flash

    verb forms that function as adjectives to describe ongoing or completed actions in sentences.

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

    This neuron activates on mentions of the subjunctive (e.g. “subjunctive,” “subjonctif,” “subjuntivo”), i.e. references to the subjunctive mood.

    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
    -switch
    -0.06
    ppe
    -0.06
    .Signal
    -0.06
     Pow
    -0.06
     seals
    -0.06
    越
    -0.06
     smoke
    -0.06
    reement
    -0.06
     agreement
    -0.06
     maxLength
    -0.06
    POSITIVE LOGITS
     NFS
    0.07
    _singular
    0.06
    angen
    0.06
    .mobile
    0.06
     conjunto
    0.06
    .red
    0.06
     назад
    0.06
    [block
    0.06
    Standard
    0.06
     uplift
    0.06
    Activations Density 0.031%

    No Known Activations