© 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. 1-RES-SM
    11. 25800
    Prev
    Next
    INDEX
    Explanations

    references to cylindrical objects or concepts related to bubbles

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    +0.19
    1.2%
    144
    +0.18
    1.0%
    43
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    144
    +0.19
    0.03
    43
    +0.18
    0.03
    367
    +0.13
    0.03
    Negative Logits
     pleasure
    -1.59
    achi
    -1.54
    wered
    -1.52
    lı
    -1.51
    ience
    -1.49
    orters
    -1.40
     Hispanic
    -1.38
    aki
    -1.35
     Sans
    -1.32
     favor
    -1.32
    POSITIVE LOGITS
    iston
    1.91
    aux
    1.74
    holder
    1.69
    head
    1.67
    ups
    1.66
    eners
    1.66
    mount
    1.64
    lite
    1.63
    heads
    1.60
    osity
    1.58
    Activations Density 0.172%

    No Known Activations