© 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
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 29692
    Prev
    Next
    INDEX
    Explanations

    events or gatherings

    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
    115
    +0.15
    0.9%
    247
    +0.13
    0.7%
    417
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    115
    +0.15
    0.02
    462
    +0.13
    0.02
    417
    +0.12
    0.02
    Negative Logits
    ¨
    -2.79
    ¹
    -2.75
    Ĥ
    -2.62
    IJ
    -2.54
    ¶
    -2.51
    ¿½
    -2.45
    ª
    -2.41
    ķ
    -2.41
    »
    -2.38
    ĸ
    -2.36
    POSITIVE LOGITS
    grounds
    2.05
    emann
    1.93
    ership
    1.87
     hostage
    1.81
     aboard
    1.75
     unanimously
    1.73
     chambers
    1.70
    away
    1.65
    held
    1.64
     annually
    1.61
    Activations Density 0.077%

    No Known Activations