© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 25830
    Prev
    Next
    INDEX
    Explanations

    various forms of questions relating to topics of inquiry

    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
    292
    +0.12
    0.6%
    479
    +0.11
    0.6%
    389
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    335
    +0.12
    0.09
    120
    +0.11
    0.09
    479
    +0.11
    0.02
    Negative Logits
    Inflater
    -1.83
     Award
    -1.56
    istes
    -1.46
    iste
    -1.45
     remainder
    -1.44
    inement
    -1.42
    ader
    -1.39
     silence
    -1.37
     Awards
    -1.36
     engagement
    -1.35
    POSITIVE LOGITS
    office
    1.67
    Credit
    1.57
    cheat
    1.54
    poons
    1.51
    lay
    1.50
    cause
    1.40
    acity
    1.40
    matter
    1.39
     drivers
    1.37
    hang
    1.37
    Activations Density 0.826%

    No Known Activations