© 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. 4-RES-SM
    11. 9371
    Prev
    Next
    INDEX
    Explanations

    references to judicial districts

    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
    +0.15
    0.9%
    78
    +0.11
    0.7%
    255
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    27
    +0.15
    0.02
    168
    +0.11
    0.02
    295
    +0.10
    0.01
    Negative Logits
    lessly
    -1.44
     precaution
    -1.39
     pars
    -1.37
    ologically
    -1.34
     conserv
    -1.32
    verts
    -1.30
     chance
    -1.25
    grown
    -1.25
    isations
    -1.23
    free
    -1.22
    POSITIVE LOGITS
     Court
    1.83
     COURT
    1.70
    Court
    1.54
    Decided
    1.47
    ÂĹ
    1.45
     Heights
    1.41
    roller
    1.41
     Magistrate
    1.41
     Attorney
    1.41
     Courts
    1.39
    Activations Density 0.029%

    No Known Activations