© 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 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. 9899
    Prev
    Next
    INDEX
    Explanations

    references to academic and scientific literature

    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
    505
    +0.13
    0.7%
    298
    +0.12
    0.7%
    365
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    365
    +0.13
    0.02
    418
    +0.12
    -0.00
    287
    +0.11
    0.02
    Negative Logits
    §
    -2.56
    ½
    -2.39
    ¿
    -2.34
    ĥ
    -2.24
    IJ
    -2.19
    Ļª
    -2.18
    »¿
    -2.17
    Ń
    -2.16
    ³
    -2.14
    Ľ
    -2.08
    POSITIVE LOGITS
    fare
    1.85
    eval
    1.84
    forum
    1.83
     literature
    1.82
    uscript
    1.81
    stud
    1.79
    Liter
    1.76
    ourse
    1.75
    catalog
    1.71
    clin
    1.68
    Activations Density 0.102%

    No Known Activations