© 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 EvalsExportsAPI 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. 12587
    Prev
    Next
    INDEX
    Explanations

    mathematical expressions and values

    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
    21
    +0.20
    1.1%
    271
    +0.17
    1.0%
    221
    +0.16
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    221
    +0.20
    0.02
    333
    +0.17
    0.02
    152
    +0.16
    0.01
    Negative Logits
    enger
    -1.65
     property
    -1.63
    engers
    -1.55
    vist
    -1.54
     actor
    -1.51
     Pradesh
    -1.46
    imals
    -1.44
     conclusions
    -1.44
    esian
    -1.43
     figure
    -1.42
    POSITIVE LOGITS
    ĨĴ
    3.84
    ĸ
    3.65
    ¶
    3.60
    3.56
    3.56
    3.56
    3.56
             ↵
    3.56
    <|outofrange|>
    3.56
                  ↵
    3.56
    Activations Density 0.060%

    No Known Activations