© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 4-RES-SM
    11. 8895
    Prev
    Next
    INDEX
    Explanations

    code structures and programming language constructs related to class and method definitions

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    307
    +0.21
    1.2%
    306
    +0.18
    1.0%
    466
    +0.15
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    17
    +0.21
    0.07
    157
    +0.18
    0.08
    78
    +0.15
    -0.01
    Negative Logits
    lying
    -1.87
    '):
    -1.77
    ]):
    -1.71
    '],
    -1.65
    >).
    -1.63
    ]),
    -1.58
    rfloor
    -1.58
    .’
    -1.56
    ’.
    -1.55
    ].)
    -1.52
    POSITIVE LOGITS
    ĥ½
    4.67
    Ģ
    4.60
      ↵
    4.58
    ↵
    4.58
    4.58
    ↵
    4.58
    4.58
    4.58
    ↵
    4.58
    ↵
    4.58
    Activations Density 0.909%

    No Known Activations