© 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. 0-MLP-SM
    11. 28248
    Prev
    Next
    INDEX
    Explanations

    the word "Which."

    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
    85
    +0.14
    0.8%
    397
    +0.12
    0.7%
    239
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    85
    +0.14
    0.02
    47
    +0.12
    0.02
    96
    +0.12
    0.02
    Negative Logits
     fiddle
    -1.60
    FIG
    -1.50
    ateness
    -1.47
    describe
    -1.39
    ![
    -1.39
    ######
    -1.36
    raising
    -1.33
     any
    -1.33
     #"
    -1.32
     learn
    -1.30
    POSITIVE LOGITS
    į
    3.55
    ĻĤ
    3.52
    Į
    3.37
    Ļª
    3.18
    ¹
    3.17
    £
    3.10
    ģ
    3.03
    ĨĴ
    3.00
    ↵
    2.98
     ↵
    2.98
    Activations Density 0.011%

    No Known Activations