© 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. 4-RES-SM
    11. 12425
    Prev
    Next
    INDEX
    Explanations

    specific legal case numbers and related judicial terminology

    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
    21
    +0.26
    1.5%
    341
    +0.14
    0.8%
    255
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    341
    +0.26
    0.02
    12
    +0.14
    0.03
    303
    +0.13
    0.03
    Negative Logits
     site
    -1.80
    irl
    -1.65
     sites
    -1.54
    atri
    -1.51
     solution
    -1.50
    ycler
    -1.50
     retrieval
    -1.47
    izi
    -1.46
     medium
    -1.45
    ään
    -1.42
    POSITIVE LOGITS
    Nos
    2.25
    Docket
    2.15
    Argued
    2.11
    UNITED
    2.06
    Appeal
    2.01
    AFFIRMED
    1.98
    NOTES
    1.89
    ¿½
    1.86
    Decided
    1.85
    Īĺ
    1.80
    Activations Density 0.053%

    No Known Activations