© 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. 12484
    Prev
    Next
    INDEX
    Explanations

    references to war crimes and legal proceedings related to them

    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
    125
    +0.13
    0.7%
    465
    +0.11
    0.6%
    54
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    465
    +0.13
    0.02
    168
    +0.11
    0.04
    496
    +0.11
    0.03
    Negative Logits
     Ravens
    -2.27
     Baltimore
    -2.24
     Kaz
    -1.87
     Ori
    -1.78
     Holmes
    -1.67
     Maryland
    -1.62
    mozilla
    -1.61
     Iz
    -1.57
    °
    -1.50
     Lew
    -1.49
    POSITIVE LOGITS
     Cambodia
    4.46
     Vietnamese
    4.42
     Thai
    4.19
     Thailand
    4.06
     Nguyen
    3.79
     Bangkok
    3.60
     Viet
    3.59
     Camb
    3.41
     Vietnam
    3.37
     Myanmar
    3.34
    Activations Density 0.938%

    No Known Activations