© 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. 31250
    Prev
    Next
    INDEX
    Explanations

    Boolean variables and their states in programming contexts

    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
    47
    +0.14
    0.8%
    82
    +0.12
    0.7%
    300
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    133
    +0.14
    0.04
    66
    +0.12
    0.04
    213
    +0.11
    0.03
    Negative Logits
    ally
    -1.79
    aling
    -1.64
    tin
    -1.63
    cerning
    -1.60
    )].
    -1.58
    inson
    -1.55
    ically
    -1.52
    APH
    -1.47
    iding
    -1.46
    cul
    -1.45
    POSITIVE LOGITS
    ³
    2.06
    ĭ
    2.06
    ģ
    1.95
     False
    1.86
    ĩ
    1.86
    Ń
    1.84
    Ĵ
    1.84
    į
    1.83
    Ļ
    1.81
    Ģ
    1.80
    Activations Density 0.143%

    No Known Activations