© 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. 30440
    Prev
    Next
    INDEX
    Explanations

    references to capacity and related metrics or constraints

    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
    118
    +0.12
    0.7%
    217
    +0.11
    0.6%
    232
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    194
    +0.12
    0.02
    415
    +0.11
    0.01
    196
    +0.10
    0.01
    Negative Logits
    anmar
    -1.78
     whom
    -1.67
    ikers
    -1.66
     ask
    -1.61
    woke
    -1.61
    ĨĴ
    -1.61
    ĩ
    -1.61
    tico
    -1.59
    ¿
    -1.58
    ±
    -1.57
    POSITIVE LOGITS
     oxidase
    1.78
    frame
    1.76
    buffer
    1.72
     capacity
    1.71
     allocated
    1.67
     range
    1.63
     ranges
    1.63
    chair
    1.61
    lif
    1.61
     structure
    1.60
    Activations Density 0.155%

    No Known Activations