© 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. 4-RES-SM
    11. 12016
    Prev
    Next
    INDEX
    Explanations

    legal terms and phrases related to court opinions and judgments

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    27
    +0.15
    0.9%
    387
    +0.13
    0.7%
    78
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    365
    +0.15
    0.04
    359
    +0.13
    0.06
    289
    +0.13
    0.03
    Negative Logits
    ries
    -1.70
     habit
    -1.50
     habits
    -1.40
     neighbourhood
    -1.40
     backgrounds
    -1.39
    ross
    -1.37
    adays
    -1.34
     interaction
    -1.32
    chaft
    -1.31
     experiments
    -1.29
    POSITIVE LOGITS
     issued
    2.05
     filed
    2.01
     thereof
    1.92
     dismissed
    1.78
    orandum
    1.72
     entered
    1.70
    filed
    1.70
     pursuant
    1.66
     denied
    1.66
     vacated
    1.66
    Activations Density 0.304%

    No Known Activations