© 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 EvalsExportsinterp-engineNEWAPI 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. 12768
    Prev
    Next
    INDEX
    Explanations

    references to speed and efficiency

    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
    156
    +0.14
    0.8%
    500
    +0.13
    0.7%
    89
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    89
    +0.14
    0.02
    361
    +0.13
    0.02
    500
    +0.12
    0.02
    Negative Logits
    thood
    -1.64
    ĻĤ
    -1.59
     pray
    -1.58
    aliana
    -1.58
    ĥ½
    -1.57
     guilt
    -1.56
     Chapel
    -1.53
     buried
    -1.49
     remanded
    -1.49
     heirs
    -1.47
    POSITIVE LOGITS
    ometer
    2.09
    ups
    2.05
    charts
    1.81
    ily
    1.79
    dial
    1.71
    imetry
    1.70
    est
    1.66
     spectrometry
    1.64
    aws
    1.64
    up
    1.59
    Activations Density 0.094%

    No Known Activations