© 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. 27805
    Prev
    Next
    INDEX
    Explanations

    dates and times in the text

    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
    153
    +0.13
    0.7%
    442
    +0.13
    0.7%
    159
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    442
    +0.13
    0.09
    408
    +0.13
    0.07
    248
    +0.12
    0.07
    Negative Logits
    į
    -2.75
    ģ
    -2.74
    ĥ
    -2.74
    ĩ
    -2.71
    §
    -2.70
    Ħ
    -2.61
    ¹
    -2.55
    Ń
    -2.54
    ħ
    -2.54
    ĸ
    -2.43
    POSITIVE LOGITS
     mornings
    1.95
     nights
    1.86
     noon
    1.83
    coming
    1.72
     hours
    1.66
     deadline
    1.62
     morning
    1.61
     night
    1.60
    cember
    1.59
    lights
    1.58
    Activations Density 2.079%

    No Known Activations