© 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. 23800
    Prev
    Next
    INDEX
    Explanations

    legal terms and concepts related to evidence and court proceedings

    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
    395
    +0.14
    0.8%
    39
    +0.12
    0.6%
    465
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    465
    +0.14
    0.12
    500
    +0.12
    0.04
    277
    +0.11
    0.11
    Negative Logits
     yourself
    -1.91
    pered
    -1.88
     yourselves
    -1.72
     trail
    -1.68
     afterward
    -1.60
     simultaneously
    -1.59
    hole
    -1.56
    .—
    -1.52
    aneous
    -1.52
     opportunities
    -1.49
    POSITIVE LOGITS
    ĥ½
    2.28
    Ļª
    1.86
    Ĩ
    1.72
    ¦
    1.68
    Ī
    1.67
    adesh
    1.64
    ´
    1.62
    ĥ
    1.58
    »¿
    1.56
    ·
    1.51
    Activations Density 2.191%

    No Known Activations