© 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. 26942
    Prev
    Next
    INDEX
    Explanations

    references to individuals' names and identities

    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
    +0.16
    0.9%
    156
    +0.15
    0.9%
    36
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    23
    +0.16
    0.04
    332
    +0.15
    0.04
    458
    +0.12
    0.03
    Negative Logits
    **]{}
    -1.85
    :**]{}
    -1.69
    ¼
    -1.69
    ”,
    -1.61
    »¿
    -1.61
    ½
    -1.58
    Ĥ¬
    -1.57
    ”—
    -1.55
    ”
    -1.55
    **]{},
    -1.54
    POSITIVE LOGITS
    hip
    2.18
    imes
    1.73
    yy
    1.68
    ially
    1.68
    hips
    1.59
    specifically
    1.58
    after
    1.52
    idades
    1.51
    rs
    1.51
    obs
    1.51
    Activations Density 0.492%

    No Known Activations