© 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. 26621
    Prev
    Next
    INDEX
    Explanations

    dialogue or quotations in text

    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
    235
    +0.12
    0.7%
    261
    +0.11
    0.6%
    216
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    147
    +0.12
    0.02
    430
    +0.11
    0.02
    509
    +0.10
    0.02
    Negative Logits
    Ķ
    -2.40
    ĥ½
    -2.02
    ¯
    -1.97
    ħ
    -1.87
    ĻĤ
    -1.86
    £
    -1.73
    Īĺ
    -1.59
    rangle
    -1.59
    İ
    -1.58
    Ĺ
    -1.54
    POSITIVE LOGITS
     humidity
    1.68
     environments
    1.67
     occurring
    1.64
     pressures
    1.58
    utes
    1.56
     temperatures
    1.53
     influences
    1.52
     asynchron
    1.48
     stresses
    1.47
     affecting
    1.47
    Activations Density 0.005%

    No Known Activations