© 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. 24167
    Prev
    Next
    INDEX
    Explanations

    the verb "to be" in its various forms, highlighting statements of existence or identity

    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
    309
    +0.12
    0.7%
    35
    +0.12
    0.7%
    279
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    63
    +0.12
    0.07
    81
    +0.12
    0.05
    499
    +0.11
    0.05
    Negative Logits
    duced
    -1.55
    gered
    -1.52
     rare
    -1.43
     ammonium
    -1.42
     bott
    -1.39
     rage
    -1.32
    import
    -1.28
     suspicion
    -1.27
    comb
    -1.26
     reuse
    -1.25
    POSITIVE LOGITS
    ¿
    1.76
    ŀ
    1.71
    oir
    1.60
    essions
    1.59
    heets
    1.57
    ĻĤ
    1.52
    lessly
    1.52
    Īĺ
    1.51
    ureus
    1.51
    .&
    1.50
    Activations Density 1.030%

    No Known Activations