© 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. 26968
    Prev
    Next
    INDEX
    Explanations

    programming constructs and syntax

    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
    74
    +0.12
    0.7%
    460
    +0.12
    0.7%
    448
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    74
    +0.12
    0.02
    131
    +0.12
    0.02
    163
    +0.11
    0.01
    Negative Logits
     chance
    -1.60
    pite
    -1.56
    hores
    -1.53
    slash
    -1.48
    $).
    -1.45
    ?).
    -1.42
    curs
    -1.41
    versus
    -1.40
    ptin
    -1.38
     deficit
    -1.37
    POSITIVE LOGITS
    ŀ
    2.23
    °
    2.21
    ¾
    2.07
    ĥ½
    2.00
    ł
    1.95
    ·
    1.93
    ¦
    1.89
    ±
    1.86
    º
    1.85
    ĻĤ
    1.85
    Activations Density 0.088%

    No Known Activations