© 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. 30934
    Prev
    Next
    INDEX
    Explanations

    concepts related to factors and conditions impacting outcomes

    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
    240
    +0.21
    1.2%
    186
    +0.16
    0.9%
    161
    +0.15
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    240
    +0.21
    0.01
    398
    +0.16
    0.10
    243
    +0.15
    0.11
    Negative Logits
    oire
    -1.92
    blogger
    -1.85
    pshire
    -1.82
    ************************************************************************
    -1.73
    oscope
    -1.63
    nai
    -1.62
    thouse
    -1.57
    gun
    -1.56
    xymatrix
    -1.54
    ³
    -1.53
    POSITIVE LOGITS
     themselves
    2.02
     differences
    1.73
     influencing
    1.72
     individually
    1.71
     combinations
    1.69
     vary
    1.69
     hormones
    1.66
     illnesses
    1.63
     confound
    1.61
     uncertainties
    1.60
    Activations Density 2.201%

    No Known Activations