© 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. 2-RES-SM
    11. 31632
    Prev
    Next
    INDEX
    Explanations

    phrases indicating capabilities or potential actions that can be taken

    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
    455
    +0.14
    0.8%
    358
    +0.12
    0.7%
    459
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    455
    +0.14
    0.10
    358
    +0.12
    0.10
    401
    +0.12
    0.07
    Negative Logits
    ¸
    -2.78
    ĻĤ
    -2.75
    Ļª
    -2.67
    ±
    -2.60
    ħ
    -2.56
    Į
    -2.52
    §
    -2.52
    ª
    -2.51
    ¹
    -2.49
    «
    -2.48
    POSITIVE LOGITS
     soon
    1.58
    forth
    1.58
    blr
    1.51
     aim
    1.46
     needed
    1.42
     myself
    1.40
    edy
    1.37
     intended
    1.36
     efficiently
    1.36
     hopefully
    1.33
    Activations Density 1.729%

    No Known Activations