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. 92944
    Prev
    Next
    INDEX
    Explanations

    English sentences

    np_max-act · gemini-2.0-flash

    The neuron activates on words that mark or start list items or numbered points (e.g. “One,” “Two,” “Using,” “For”).

    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
    ?).
    -0.08
    ’.
    -0.07
    '.
    -0.07
    )).
    -0.06
    ”.
    -0.06
    	sp
    -0.06
     ~/.
    -0.06
     checkBox
    -0.06
     ).
    -0.06
     ];
    -0.06
    POSITIVE LOGITS
    Throughout
    0.06
    _countries
    0.06
     přen
    0.06
    есь
    0.06
    Naz
    0.06
     NaN
    0.06
                
    ↵
    0.06
     oyun
    0.06
    ises
    0.06
    _track
    0.06
    Activations Density 0.450%

    No Known Activations