© 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. Residuals
    9. 1-RES-SM
    11. 26688
    Prev
    Next
    INDEX
    Explanations

    references to types of "kits" or categories in general

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__res-sm_processed/1-res-sm
    Prompts (Dashboard)
    32,768 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.1.hook_resid_post
    Hook Layer
    1
    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
    18
    +0.11
    0.7%
    466
    +0.11
    0.7%
    249
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    466
    +0.11
    0.01
    217
    +0.11
    0.01
    41
    +0.10
    0.01
    Negative Logits
    odot
    -1.71
     sight
    -1.65
     studied
    -1.57
    itance
    -1.50
     else
    -1.49
     polymorphic
    -1.43
     pronounced
    -1.42
     crowded
    -1.40
     sudden
    -1.39
    ilate
    -1.39
    POSITIVE LOGITS
    kowski
    1.93
    eries
    1.71
    coin
    1.70
    ño
    1.69
    emen
    1.55
    âĢIJ
    1.55
    aho
    1.52
    ing
    1.48
    ious
    1.47
     presses
    1.47
    Activations Density 0.145%

    No Known Activations