© 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. 23432
    Prev
    Next
    INDEX
    Explanations

    terms related to data transmission and management in networking

    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
    +0.15
    0.9%
    358
    +0.13
    0.7%
    229
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    120
    +0.15
    0.04
    311
    +0.13
    0.04
    127
    +0.12
    0.04
    Negative Logits
     degrees
    -1.86
     missions
    -1.74
     concerts
    -1.64
     generation
    -1.60
     respective
    -1.59
     etc
    -1.53
     pes
    -1.52
     caliber
    -1.50
     suits
    -1.50
     contributions
    -1.50
    POSITIVE LOGITS
     ':
    1.77
    expected
    1.65
     '</
    1.64
    assertEqual
    1.61
     '#
    1.58
     "</
    1.57
    assert
    1.55
    ĥ½
    1.53
     _"
    1.51
    assertEquals
    1.51
    Activations Density 0.291%

    No Known Activations