© 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. 2-RES-SM
    11. 13326
    Prev
    Next
    INDEX
    Explanations

    patterns that indicate complex mathematical or technical relationships

    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/2-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.2.hook_resid_post
    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
    204
    +0.15
    0.9%
    198
    +0.12
    0.7%
    72
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    204
    +0.15
    0.14
    72
    +0.12
    0.09
    398
    +0.12
    0.12
    Negative Logits
     revocation
    -1.51
    uance
    -1.49
    opolys
    -1.42
    '>
    -1.37
    âĢIJ
    -1.32
     issuance
    -1.32
     matter
    -1.31
     creation
    -1.30
     SSH
    -1.29
     transmitted
    -1.29
    POSITIVE LOGITS
    ľĵ
    3.10
    ¥
    3.04
    ŀ
    2.95
    ľ
    2.92
    ¼
    2.89
    ¤
    2.87
    Ī
    2.83
    ĻĤ
    2.83
    Ĵ
    2.83
    °
    2.79
    Activations Density 6.017%

    No Known Activations