© 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. 26654
    Prev
    Next
    INDEX
    Explanations

    elements related to gradients and color codes in CSS

    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
    263
    +0.19
    1.1%
    56
    +0.13
    0.8%
    320
    +0.11
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    191
    +0.19
    0.13
    323
    +0.13
    0.04
    263
    +0.11
    0.15
    Negative Logits
    Decided
    -1.49
    ecycle
    -1.40
     citations
    -1.36
    ibraries
    -1.35
    etheus
    -1.35
     Papers
    -1.34
    :</
    -1.33
     credit
    -1.31
     voltages
    -1.30
    SHIFT
    -1.30
    POSITIVE LOGITS
    uble
    1.45
     guaranteed
    1.35
     ranch
    1.34
    ener
    1.33
    asone
    1.27
     son
    1.22
     safe
    1.22
     Airport
    1.22
    urane
    1.21
     protected
    1.21
    Activations Density 5.000%

    No Known Activations