© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 29691
    Prev
    Next
    INDEX
    Explanations

    references to locations and numbers

    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
    410
    +0.13
    0.7%
    156
    +0.13
    0.7%
    183
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    165
    +0.13
    0.02
    202
    +0.13
    0.02
    183
    +0.12
    0.02
    Negative Logits
    àµį
    -2.09
    watson
    -1.71
    à±į
    -1.63
    uximab
    -1.62
    tons
    -1.62
    egen
    -1.57
    unov
    -1.57
    etry
    -1.54
    cex
    -1.54
    à±ģ
    -1.54
    POSITIVE LOGITS
    ·
    2.45
    ¿½
    2.37
    IJ
    2.23
    Ŀ
    2.01
    ¸
    1.95
    ³
    1.94
    §
    1.93
    ĸ´
    1.90
    ī
    1.86
    Ĭ
    1.83
    Activations Density 0.041%

    No Known Activations