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

    Neuronpedia

    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. MLP Post
    9. 0-MLP-SM
    11. 28092
    Prev
    Next
    INDEX
    Explanations

    instances where the word "only" is used to emphasize exclusivity or limitation

    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
    128
    +0.12
    0.7%
    201
    +0.12
    0.6%
    409
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    128
    +0.12
    0.09
    409
    +0.12
    0.09
    207
    +0.12
    0.09
    Negative Logits
    matic
    -1.88
    ampoline
    -1.51
    rVert
    -1.51
     neither
    -1.50
    ulic
    -1.46
    rvert
    -1.46
    ment
    -1.46
    inates
    -1.44
    inating
    -1.41
    inct
    -1.38
    POSITIVE LOGITS
     complaints
    1.47
    reb
    1.45
     novels
    1.43
     traged
    1.41
    veh
    1.41
    ournal
    1.39
     conceivable
    1.38
     publications
    1.37
     miss
    1.36
    öl
    1.36
    Activations Density 5.068%

    No Known Activations