© 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 EvalsExportsinterp-engineNEWAPI 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. 22819
    Prev
    Next
    INDEX
    Explanations

    articles and their associated variations in usage

    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
    291
    +0.12
    0.7%
    196
    +0.11
    0.6%
    217
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    496
    +0.12
    0.05
    291
    +0.11
    0.06
    480
    +0.11
    0.04
    Negative Logits
     Americans
    -1.56
     Basic
    -1.47
     used
    -1.36
     bile
    -1.35
     graft
    -1.31
     blood
    -1.31
     perox
    -1.29
    lay
    -1.29
    Âķ
    -1.28
     Applied
    -1.28
    POSITIVE LOGITS
    ozy
    1.91
    /)
    1.74
    ific
    1.71
    ographically
    1.66
    iste
    1.65
    itatively
    1.61
     contrary
    1.58
    urally
    1.58
    istically
    1.58
    entially
    1.58
    Activations Density 0.628%

    No Known Activations