© 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. 13961
    Prev
    Next
    INDEX
    Explanations

    references to "Real" in different contexts, indicating a focus on elements related to reality or realism

    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
    507
    +0.13
    0.7%
    338
    +0.12
    0.7%
    20
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    85
    +0.13
    0.02
    338
    +0.12
    0.01
    30
    +0.11
    0.01
    Negative Logits
    urable
    -1.69
    hene
    -1.47
     downstream
    -1.43
    ively
    -1.42
    iewicz
    -1.40
     href
    -1.38
    ivity
    -1.36
    else
    -1.36
    aque
    -1.33
    urdy
    -1.32
    POSITIVE LOGITS
    ĺ
    3.27
    ģ
    3.08
    ĻĤ
    2.95
    ĥ½
    2.94
    ¾
    2.82
    ¼
    2.81
    ī
    2.81
    ½
    2.79
    ł
    2.77
    Ľ
    2.77
    Activations Density 0.174%

    No Known Activations