© 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. 12849
    Prev
    Next
    INDEX
    Explanations

    the use of mathematical symbols and notation in equations

    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
    241
    +0.16
    0.9%
    78
    +0.14
    0.8%
    426
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    241
    +0.16
    0.02
    129
    +0.14
    0.01
    166
    +0.13
    0.01
    Negative Logits
     altogether
    -1.68
    ter
    -1.61
    nel
    -1.51
    forth
    -1.51
    ubicin
    -1.50
    ized
    -1.48
    ically
    -1.44
    gi
    -1.41
    aster
    -1.39
    nam
    -1.39
    POSITIVE LOGITS
    ı
    3.14
    ŀ
    3.03
    ³
    2.99
    «
    2.97
    ¶
    2.92
    §
    2.91
    °
    2.88
    Ŀ
    2.86
    Ĵ
    2.79
    ´
    2.71
    Activations Density 0.076%

    No Known Activations