© 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. 10348
    Prev
    Next
    INDEX
    Explanations

    questions and answers formatted in a structured or computational style

    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
    505
    +0.12
    0.7%
    425
    +0.11
    0.6%
    302
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    505
    +0.12
    0.02
    91
    +0.11
    0.02
    289
    +0.11
    0.01
    Negative Logits
    WARE
    -1.58
    uchs
    -1.49
    workers
    -1.43
    IGEN
    -1.40
    ker
    -1.36
    sk
    -1.35
    cal
    -1.35
     vet
    -1.35
    oles
    -1.33
    cker
    -1.30
    POSITIVE LOGITS
    ī
    2.10
    »
    1.89
    ¼
    1.79
    ĥ½
    1.71
    ¥
    1.69
    µ
    1.67
    icles
    1.65
    Ī
    1.58
     intercepted
    1.55
    »¿
    1.54
    Activations Density 0.065%

    No Known Activations