© 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. 12679
    Prev
    Next
    INDEX
    Explanations

    structural elements and syntactical components in code and data

    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
    419
    +0.23
    1.4%
    353
    +0.13
    0.8%
    12
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    21
    +0.23
    0.09
    419
    +0.13
    0.10
    143
    +0.12
    0.07
    Negative Logits
    ship
    -1.62
    irl
    -1.48
    ition
    -1.47
    ships
    -1.46
    reek
    -1.40
    izens
    -1.37
    rais
    -1.35
    ols
    -1.34
    icals
    -1.33
    rapeutics
    -1.32
    POSITIVE LOGITS
    ¾
    1.62
    }}$-
    1.58
    Ŀ
    1.56
     sides
    1.55
    anian
    1.47
    ī
    1.41
    ŀ
    1.41
    ħ
    1.39
    ÏĥηÏĤ
    1.38
    ±
    1.36
    Activations Density 3.635%

    No Known Activations