© 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. 23362
    Prev
    Next
    INDEX
    Explanations

    phrases related to experimental procedures and scientific results

    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
    156
    +0.19
    1.1%
    170
    +0.14
    0.8%
    347
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    170
    +0.19
    0.03
    410
    +0.14
    0.02
    238
    +0.13
    0.01
    Negative Logits
    \]](
    -1.72
    ately
    -1.51
    ally
    -1.51
     thanks
    -1.48
     controversy
    -1.46
     approximately
    -1.46
     Syndrome
    -1.40
    arily
    -1.40
    Argued
    -1.39
    ats
    -1.38
    POSITIVE LOGITS
    vez
    1.73
     liter
    1.48
    gression
    1.46
    zos
    1.44
     favor
    1.35
    thems
    1.35
     Investment
    1.35
    encial
    1.34
     heirs
    1.34
     ún
    1.33
    Activations Density 0.266%

    No Known Activations