© 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. 31870
    Prev
    Next
    INDEX
    Explanations

    references to medical terminology and processes

    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
    369
    +0.18
    1.0%
    497
    +0.14
    0.8%
    74
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    103
    +0.18
    0.04
    74
    +0.14
    0.03
    45
    +0.14
    0.02
    Negative Logits
    except
    -1.61
    ante
    -1.53
    tion
    -1.51
    ever
    -1.45
     except
    -1.42
    soever
    -1.41
    edes
    -1.39
    able
    -1.39
    ()"
    -1.37
     winning
    -1.35
    POSITIVE LOGITS
    ¡
    2.66
    ¦
    2.63
    ĥ½
    2.58
    ¬
    2.47
    ģ
    2.44
    ĨĴ
    2.40
    ľ
    2.35
    ŀ
    2.29
    µ
    2.28
    ±
    2.27
    Activations Density 0.162%

    No Known Activations