© 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. 26739
    Prev
    Next
    INDEX
    Explanations

    quotes or phrases that indicate references or citations within a document

    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
    156
    +0.20
    1.1%
    422
    +0.12
    0.7%
    479
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    469
    +0.20
    0.09
    7
    +0.12
    0.00
    451
    +0.12
    0.04
    Negative Logits
    ľĵ
    -1.50
    ses
    -1.48
    IJ
    -1.45
     thrive
    -1.43
    arman
    -1.40
    umni
    -1.39
    uttering
    -1.39
    ires
    -1.38
    CDATA
    -1.38
    allas
    -1.38
    POSITIVE LOGITS
     proceeds
    1.62
     dock
    1.37
     anhydr
    1.36
     supra
    1.31
     sight
    1.29
     Kitchen
    1.25
    Layer
    1.22
    chitz
    1.21
    acin
    1.21
     Stick
    1.20
    Activations Density 3.434%

    No Known Activations