© 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. 12689
    Prev
    Next
    INDEX
    Explanations

    phrases indicating impossibility or restrictions

    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
    358
    +0.16
    0.9%
    415
    +0.14
    0.8%
    202
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    500
    +0.16
    0.05
    358
    +0.14
    0.05
    415
    +0.11
    0.03
    Negative Logits
     años
    -1.47
    edom
    -1.46
    ி
    -1.40
     degrees
    -1.39
    CTIONS
    -1.38
     differently
    -1.38
    ulf
    -1.37
     millenn
    -1.36
     challenges
    -1.36
     Civil
    -1.35
    POSITIVE LOGITS
     anymore
    1.95
    yet
    1.82
     whatsoever
    1.69
    wegian
    1.69
     yet
    1.68
    oriously
    1.63
     ever
    1.63
     exceed
    1.60
    forge
    1.56
    exist
    1.56
    Activations Density 0.247%

    No Known Activations