© 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. 13413
    Prev
    Next
    INDEX
    Explanations

    phrases related to a sequence of events or topics presented in a series format

    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
    33
    +0.12
    0.7%
    393
    +0.12
    0.7%
    343
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    511
    +0.12
    0.01
    343
    +0.12
    0.02
    68
    +0.11
    0.02
    Negative Logits
    ľĵ
    -1.81
    IJ
    -1.76
    ŀ
    -1.72
    ĥ
    -1.71
    §
    -1.68
    ĥ½
    -1.66
    Ń
    -1.63
     fertile
    -1.61
    ½
    -1.56
    ¸
    -1.55
    POSITIVE LOGITS
     expansions
    2.05
    inator
    2.02
     finale
    1.94
    ional
    1.90
     expansion
    1.85
    helf
    1.83
    olutions
    1.75
    atics
    1.74
    olar
    1.71
    ematic
    1.70
    Activations Density 0.119%

    No Known Activations