© 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. 13655
    Prev
    Next
    INDEX
    Explanations

    numerical values and statistical 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
    326
    +0.19
    1.1%
    160
    +0.15
    0.9%
    186
    +0.15
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    160
    +0.19
    0.17
    256
    +0.15
    0.15
    326
    +0.15
    0.07
    Negative Logits
    Ļ
    -1.82
    openh
    -1.61
     comment
    -1.44
    eca
    -1.39
     options
    -1.35
     notes
    -1.35
     reminds
    -1.35
     alternatives
    -1.33
     Chart
    -1.32
     stainless
    -1.31
    POSITIVE LOGITS
    bid
    1.55
    forced
    1.48
    Major
    1.47
    Reuters
    1.46
    fore
    1.44
    Unknown
    1.42
    Secondary
    1.41
    ![**
    1.40
    '"
    1.39
    apon
    1.39
    Activations Density 4.830%

    No Known Activations