© 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. 2-RES-SM
    11. 12555
    Prev
    Next
    INDEX
    Explanations

    references to bridges and their features

    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/2-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.2.hook_resid_post
    Hook Layer
    2
    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
    43
    +0.17
    1.0%
    156
    +0.16
    0.9%
    304
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    304
    +0.17
    0.03
    43
    +0.16
    0.02
    443
    +0.14
    0.02
    Negative Logits
    iasis
    -1.70
    opic
    -1.63
    ureus
    -1.60
     favor
    -1.52
     odds
    -1.50
     fetus
    -1.47
     Zeit
    -1.46
     little
    -1.46
     vez
    -1.45
    roph
    -1.44
    POSITIVE LOGITS
    lands
    2.04
    bows
    1.89
     towers
    1.81
    roads
    1.79
     architecture
    1.79
     tower
    1.72
    mounted
    1.70
    yard
    1.67
     arches
    1.66
     bridges
    1.64
    Activations Density 0.240%

    No Known Activations