© 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. 2-RES-SM
    11. 28875
    Prev
    Next
    INDEX
    Explanations

    terms related to conversion and conversion tools

    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
    156
    +0.14
    0.8%
    148
    +0.13
    0.8%
    376
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    148
    +0.14
    0.01
    62
    +0.13
    0.01
    474
    +0.12
    0.01
    Negative Logits
    fare
    -2.10
    iento
    -1.79
    hood
    -1.67
    lies
    -1.60
     else
    -1.58
    lessness
    -1.58
    rens
    -1.50
    kill
    -1.45
    friends
    -1.45
     dissection
    -1.45
    POSITIVE LOGITS
    itive
    1.85
    esium
    1.70
    herent
    1.70
    ivated
    1.60
    atable
    1.60
     thereto
    1.58
    ipher
    1.49
    ational
    1.47
    elesc
    1.44
     converts
    1.44
    Activations Density 0.125%

    No Known Activations