© 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. 11390
    Prev
    Next
    INDEX
    Explanations

    references to prior art or previous instances in various contexts

    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
    21
    +0.20
    1.2%
    11
    +0.13
    0.8%
    270
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    11
    +0.20
    0.01
    487
    +0.13
    0.01
    18
    +0.11
    0.01
    Negative Logits
    «
    -1.68
    Ļ
    -1.57
    £
    -1.52
    ional
    -1.52
    omycin
    -1.50
    ľĵ
    -1.45
    ¾
    -1.45
    while
    -1.44
    uximab
    -1.42
    ĭ
    -1.37
    POSITIVE LOGITS
     Invention
    2.13
     Art
    2.02
     invention
    1.80
     Publication
    1.75
     inventions
    1.72
     compositions
    1.72
     art
    1.71
     Arts
    1.65
    ties
    1.64
     thereto
    1.63
    Activations Density 0.013%

    No Known Activations