© 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. 23024
    Prev
    Next
    INDEX
    Explanations

    conjunctions connecting phrases or ideas

    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
    371
    +0.11
    0.6%
    102
    +0.10
    0.5%
    107
    +0.09
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    452
    +0.11
    0.05
    102
    +0.10
    0.05
    210
    +0.09
    0.02
    Negative Logits
    .);
    -1.64
    .).
    -1.62
    liv
    -1.57
    .),
    -1.56
    Simplify
    -1.53
    versed
    -1.49
     nerv
    -1.47
    pper
    -1.43
    ellants
    -1.42
    iph
    -1.40
    POSITIVE LOGITS
    ī
    3.45
    ģ
    3.09
    Ĵ
    3.06
    ĭ
    2.87
    ³
    2.85
    Ģ
    2.79
    ·
    2.78
    ª
    2.78
    Ŀ
    2.77
    »
    2.75
    Activations Density 0.582%

    No Known Activations