© 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. Residuals
    9. 1-RES-SM
    11. 29464
    Prev
    Next
    INDEX
    Explanations

    abbreviations or acronyms related to medical terms or conditions

    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
    369
    +0.20
    1.1%
    39
    +0.14
    0.8%
    192
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    369
    +0.20
    0.05
    39
    +0.14
    0.05
    412
    +0.12
    0.05
    Negative Logits
     METHODS
    -1.64
    chy
    -1.61
    ties
    -1.57
    ests
    -1.55
    ching
    -1.55
    ateurs
    -1.53
    TRODUCTION
    -1.52
     occasions
    -1.47
    ders
    -1.43
    king
    -1.39
    POSITIVE LOGITS
    Ļª
    2.69
    ŀ
    2.68
    Ŀ
    2.62
    ·
    2.54
    Ĵ
    2.38
    ®
    2.30
    Į
    2.22
    ĺ
    2.20
    ³
    2.20
    Ī
    2.18
    Activations Density 0.179%

    No Known Activations