© 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. 2-RES-SM
    11. 9724
    Prev
    Next
    INDEX
    Explanations

    phrases and terms related to rankings and best-in-category references

    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
    457
    +0.17
    0.9%
    118
    +0.13
    0.7%
    305
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    457
    +0.17
    0.04
    305
    +0.13
    0.02
    118
    +0.11
    0.04
    Negative Logits
    ually
    -1.98
    ANCE
    -1.68
    utable
    -1.63
    istered
    -1.59
    umber
    -1.59
    ibly
    -1.56
    usement
    -1.55
     herein
    -1.55
    istration
    -1.52
    vity
    -1.49
    POSITIVE LOGITS
    ographical
    2.17
     notch
    2.10
    ographies
    2.02
     ranked
    1.93
     tier
    1.80
     drawer
    1.77
     rated
    1.73
    ellt
    1.66
    icked
    1.64
    het
    1.64
    Activations Density 0.102%

    No Known Activations