© 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. 24844
    Prev
    Next
    INDEX
    Explanations

    patterns related to programming or scripting languages and their elements

    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
    23
    +0.25
    1.4%
    493
    +0.13
    0.7%
    484
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    23
    +0.25
    0.06
    424
    +0.13
    0.04
    484
    +0.11
    0.01
    Negative Logits
     públic
    -1.76
     remedies
    -1.70
     magazine
    -1.64
     financing
    -1.62
    ismo
    -1.62
     objectives
    -1.61
    writer
    -1.58
     forum
    -1.57
     promotion
    -1.55
    azine
    -1.52
    POSITIVE LOGITS
     older
    1.83
    semb
    1.58
    lder
    1.57
    >$
    1.56
     none
    1.49
    chers
    1.47
    ]{.
    1.47
     elderly
    1.46
    OCT
    1.45
     resemble
    1.45
    Activations Density 1.691%

    No Known Activations