© 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. 30809
    Prev
    Next
    INDEX
    Explanations

    technical terminology and programming constructs related to software or web development

    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
    263
    +0.20
    1.1%
    286
    +0.12
    0.7%
    400
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    263
    +0.20
    0.17
    44
    +0.12
    0.09
    340
    +0.11
    0.08
    Negative Logits
     instant
    -1.43
     genus
    -1.38
    erty
    -1.33
     undertaking
    -1.31
     elegans
    -1.29
    orate
    -1.29
    landers
    -1.28
    idable
    -1.27
     appreciate
    -1.27
     Liberation
    -1.26
    POSITIVE LOGITS
    ľĵ
    3.40
    ĻĤ
    3.23
    ®
    3.02
    ĨĴ
    3.01
    Ļª
    2.85
    Ĥ¬
    2.85
    ¦
    2.83
    ĭ
    2.82
    Ĥ
    2.82
    ĥ½
    2.79
    Activations Density 5.436%

    No Known Activations