© 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. 1-RES-SM
    11. 25334
    Prev
    Next
    INDEX
    Explanations

    the repeated mention of the term "Mon."

    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
    369
    +0.21
    1.2%
    477
    +0.13
    0.8%
    59
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    477
    +0.21
    0.03
    59
    +0.13
    0.02
    391
    +0.13
    0.02
    Negative Logits
     poor
    -1.74
    lass
    -1.70
    ESULT
    -1.63
     authenticity
    -1.51
     diagnosis
    -1.49
     limited
    -1.46
    effectiveness
    -1.42
    rapeutic
    -1.40
     obsc
    -1.40
     prognosis
    -1.36
    POSITIVE LOGITS
    ettes
    2.18
    ths
    2.11
    ette
    1.89
    stadt
    1.74
    nes
    1.74
    atha
    1.64
    ett
    1.63
    quette
    1.61
    sey
    1.57
    zes
    1.56
    Activations Density 0.050%

    No Known Activations