© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. MLP Post
    9. 0-MLP-SM
    11. 28431
    Prev
    Next
    INDEX
    Explanations

    references to plural forms of nouns

    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/0-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.0.hook_mlp_out
    Hook Layer
    0
    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
    413
    +0.10
    0.6%
    257
    +0.10
    0.6%
    15
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    413
    +0.10
    0.01
    257
    +0.10
    0.01
    121
    +0.10
    0.01
    Negative Logits
    hes
    -1.90
    rian
    -1.64
    rogen
    -1.64
     Eve
    -1.64
    rogens
    -1.60
    roc
    -1.51
     close
    -1.48
    reign
    -1.42
    ther
    -1.40
     anybody
    -1.39
    POSITIVE LOGITS
    Ĭ
    2.73
    £
    2.73
    ¢
    2.69
    ¼
    2.59
    Ł
    2.54
    ¾
    2.51
    ĩ
    2.44
    º
    2.42
    ĸ
    2.39
    ĵ
    2.37
    Activations Density 0.010%

    No Known Activations