© 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. MLP Post
    9. 2-MLP-SM
    11. 24546
    Prev
    Next
    INDEX
    Explanations

    phrases describing optimal methods or approaches for achieving goals

    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/2-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.2.hook_mlp_out
    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
    127
    +0.13
    0.7%
    439
    +0.12
    0.7%
    223
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    372
    +0.13
    0.03
    223
    +0.12
    0.03
    127
    +0.12
    0.04
    Negative Logits
    APTER
    -1.52
    powers
    -1.50
     luck
    -1.48
    documentclass
    -1.48
    issions
    -1.47
     heels
    -1.43
    ubuntu
    -1.38
    tails
    -1.36
     talents
    -1.34
    KS
    -1.30
    POSITIVE LOGITS
    udo
    1.58
    feld
    1.54
    liness
    1.51
     than
    1.51
    obox
    1.50
    uria
    1.48
    atically
    1.46
    osti
    1.46
    meric
    1.37
    lio
    1.37
    Activations Density 0.105%

    No Known Activations