© 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. 25216
    Prev
    Next
    INDEX
    Explanations

    instances of the word "began."

    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
    376
    +0.19
    1.1%
    43
    +0.17
    1.0%
    317
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    43
    +0.19
    0.04
    167
    +0.17
    0.03
    106
    +0.13
    0.03
    Negative Logits
    adem
    -1.51
    erne
    -1.49
    kte
    -1.49
    asses
    -1.47
    opard
    -1.47
    keywords
    -1.43
    figure
    -1.42
    uten
    -1.36
    ounded
    -1.36
    case
    -1.35
    POSITIVE LOGITS
     WARRANTY
    1.76
    obar
    1.54
     adulthood
    1.45
     hereto
    1.42
    ogonal
    1.38
     innoc
    1.38
    above
    1.35
    onymous
    1.34
     correctly
    1.30
     missing
    1.30
    Activations Density 0.022%

    No Known Activations