© 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. 13394
    Prev
    Next
    INDEX
    Explanations

    legal terminology and references to law or 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/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
    447
    +0.14
    0.8%
    21
    +0.14
    0.8%
    334
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    450
    +0.14
    0.02
    64
    +0.14
    0.06
    255
    +0.13
    -0.01
    Negative Logits
    ated
    -1.80
    ¡
    -1.78
    ates
    -1.77
    ľĵ
    -1.74
    atory
    -1.71
    atically
    -1.70
    ically
    -1.67
    ¶
    -1.64
    ĻĤ
    -1.61
    ially
    -1.60
    POSITIVE LOGITS
     fame
    1.55
    ld
    1.37
    lem
    1.32
    ãģĪ
    1.31
    ras
    1.30
    екÑĤ
    1.26
    liga
    1.26
    beit
    1.25
    umi
    1.25
     ago
    1.25
    Activations Density 0.605%

    No Known Activations