© 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
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 2-RES-SM
    11. 12300
    Prev
    Next
    INDEX
    Explanations

    references to legal actions and court proceedings

    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
    111
    +0.24
    1.4%
    156
    +0.20
    1.2%
    34
    +0.11
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    410
    +0.24
    0.11
    266
    +0.20
    0.06
    63
    +0.11
    0.06
    Negative Logits
    gor
    -1.64
    fered
    -1.54
    ftware
    -1.47
    ppa
    -1.44
     [-
    -1.42
     [
    -1.37
    ness
    -1.36
    GA
    -1.34
    bol
    -1.30
    representation
    -1.30
    POSITIVE LOGITS
    »¿
    4.25
    ĻĤ
    4.19
    Īĺ
    3.80
    ¡
    3.64
    ĸ´
    3.62
    ³
    3.60
    ²
    3.58
    ¼
    3.57
    ĺ
    3.51
    Ļ
    3.44
    Activations Density 1.894%

    No Known Activations