© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 10786
    Prev
    Next
    INDEX
    Explanations

    copyright information and licensing details

    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
    263
    +0.14
    0.8%
    385
    +0.12
    0.7%
    448
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    74
    +0.14
    0.04
    192
    +0.12
    0.03
    107
    +0.10
    0.04
    Negative Logits
    Inflater
    -1.67
    ullary
    -1.63
     voltage
    -1.47
     occurs
    -1.41
    $.[]{
    -1.39
    gia
    -1.37
    reptococcus
    -1.36
    .[]{
    -1.36
    /-/
    -1.33
    monella
    -1.32
    POSITIVE LOGITS
     Begin
    1.62
     Teams
    1.57
     Hour
    1.53
     teams
    1.46
    aming
    1.45
     Things
    1.40
    seeing
    1.37
    amed
    1.31
     Cookies
    1.31
     meetings
    1.29
    Activations Density 0.318%

    No Known Activations