© 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. 31262
    Prev
    Next
    INDEX
    Explanations

    terms and numerical values related to measurements and scientific data

    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
    157
    +0.14
    0.8%
    326
    +0.14
    0.8%
    102
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    74
    +0.14
    0.02
    315
    +0.14
    0.01
    391
    +0.12
    0.02
    Negative Logits
    after
    -1.50
     outset
    -1.50
    ause
    -1.47
    ollary
    -1.42
     latter
    -1.41
     reviewer
    -1.39
    tikz
    -1.36
    arin
    -1.35
     quo
    -1.33
     venue
    -1.32
    POSITIVE LOGITS
     kilograms
    1.52
     refriger
    1.46
     Array
    1.44
     milliseconds
    1.44
     micrograms
    1.43
    00000000
    1.42
    >`
    1.39
    imum
    1.35
    }</
    1.35
    kil
    1.32
    Activations Density 0.093%

    No Known Activations