© 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. 13769
    Prev
    Next
    INDEX
    Explanations

    variables and their types in programming or coding 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
    122
    +0.18
    1.0%
    423
    +0.15
    0.9%
    342
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    122
    +0.18
    0.09
    450
    +0.15
    0.07
    168
    +0.13
    0.05
    Negative Logits
    tat
    -1.74
    pective
    -1.65
    ija
    -1.64
    tawa
    -1.58
    rification
    -1.48
    velopment
    -1.43
    rus
    -1.40
    rosis
    -1.36
    kerchief
    -1.35
    ousand
    -1.34
    POSITIVE LOGITS
     grounds
    1.45
     herself
    1.36
     myself
    1.35
    ť
    1.34
     Natural
    1.32
     ourselves
    1.31
     Intelligence
    1.30
    cke
    1.26
    CASE
    1.24
     Eve
    1.23
    Activations Density 1.927%

    No Known Activations