© 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. 12591
    Prev
    Next
    INDEX
    Explanations

    terms related to connectivity and configuration in systems or devices

    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
    348
    +0.24
    1.5%
    78
    +0.13
    0.8%
    469
    +0.09
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    348
    +0.24
    0.02
    352
    +0.13
    0.01
    289
    +0.09
    0.01
    Negative Logits
    º
    -1.36
    woke
    -1.31
     deepest
    -1.30
    ¿
    -1.29
    cession
    -1.25
    quiry
    -1.23
     recursion
    -1.20
     guts
    -1.17
     cliff
    -1.17
     lag
    -1.16
    POSITIVE LOGITS
     thereto
    3.03
     therein
    2.55
     thereon
    2.48
    xe
    2.33
     thereof
    2.31
     herein
    2.22
     therefrom
    2.17
     Patent
    2.05
     hereto
    1.99
     wherein
    1.95
    Activations Density 0.090%

    No Known Activations