© 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. 13139
    Prev
    Next
    INDEX
    Explanations

    terms related to internships and their associated terms

    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
    343
    +0.16
    0.9%
    59
    +0.12
    0.7%
    494
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    59
    +0.16
    0.05
    200
    +0.12
    0.06
    343
    +0.12
    0.07
    Negative Logits
     luck
    -1.75
    wn
    -1.60
     nothing
    -1.57
    arma
    -1.45
    ://
    -1.45
     sender
    -1.43
    None
    -1.36
    rote
    -1.30
     behalf
    -1.30
    pped
    -1.29
    POSITIVE LOGITS
    ©
    3.88
    ¨
    3.84
    ±
    3.79
    ³
    3.60
    İ
    3.58
    ª
    3.54
    Ĭ
    3.46
    ¬
    3.41
    ¿
    3.40
    ¼
    3.40
    Activations Density 0.961%

    No Known Activations