© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 2-RES-SM
    11. 29638
    Prev
    Next
    INDEX
    Explanations

    references to tables or data structures

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__res-sm_processed/2-res-sm
    Prompts (Dashboard)
    32,768 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.2.hook_resid_post
    Hook Layer
    2
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    243
    +0.15
    0.8%
    224
    +0.13
    0.7%
    357
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    357
    +0.15
    0.01
    79
    +0.13
    0.01
    224
    +0.12
    0.01
    Negative Logits
     charge
    -1.79
    ness
    -1.72
     harmless
    -1.67
    Ļª
    -1.66
    «
    -1.66
    Ń
    -1.62
    ĸ
    -1.61
    ¬
    -1.55
    ½
    -1.55
    ¥
    -1.54
    POSITIVE LOGITS
    au
    1.94
    bed
    1.89
    top
    1.86
     Tenn
    1.86
    mania
    1.83
     tennis
    1.81
    cloth
    1.78
    backs
    1.61
     Rum
    1.61
    rett
    1.59
    Activations Density 0.057%

    No Known Activations