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. 20585
    Prev
    Next
    INDEX
    Explanations

    ogenic

    np_max-act · gemini-2.0-flash

    The neuron activates on technical terms that denote cancer‐causing or mutagenic properties (i.e. references to carcinogenicity and related toxicology).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    terms related to carcinogenicity and cancer mechanisms.

    oai_token-act-pair · gpt-4o-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
     Checkout
    -0.06
     commentators
    -0.06
     separates
    -0.06
     nhằm
    -0.06
    -blue
    -0.06
     questioning
    -0.06
     outreach
    -0.06
     trouvé
    -0.06
     opacity
    -0.06
    _validator
    -0.06
    POSITIVE LOGITS
     získ
    0.07
    	transform
    0.07
     nihil
    0.06
    _piece
    0.06
    ffer
    0.06
     harmless
    0.06
    pc
    0.06
    $string
    0.06
    osate
    0.06
     chained
    0.06
    Activations Density 0.006%

    No Known Activations