© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. MLP Post
    9. 0-MLP-SM
    11. 25221
    Prev
    Next
    INDEX
    Explanations

    references to visual data representations such as charts and graphs

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__mlp-sm_processed/0-mlp-sm
    Prompts (Dashboard)
    32,768 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.0.hook_mlp_out
    Hook Layer
    0
    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
    19
    +0.12
    0.7%
    244
    +0.11
    0.7%
    109
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    19
    +0.12
    0.01
    293
    +0.11
    0.01
    244
    +0.10
    0.01
    Negative Logits
    rians
    -1.85
     privilege
    -1.83
     grateful
    -1.73
     monol
    -1.72
     privileged
    -1.72
     disbelief
    -1.69
     privileges
    -1.66
     happier
    -1.63
     fear
    -1.61
     sexes
    -1.59
    POSITIVE LOGITS
    heet
    2.33
    helf
    2.31
    imen
    1.97
    ball
    1.93
    ource
    1.93
    mith
    1.93
    doctoral
    1.83
    iewicz
    1.82
    piece
    1.76
    oon
    1.74
    Activations Density 0.040%

    No Known Activations