© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 30149
    Prev
    Next
    INDEX
    Explanations

    references to neural mechanisms and their mediation processes

    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
    +0.15
    0.9%
    419
    +0.15
    0.8%
    500
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    149
    +0.15
    0.08
    240
    +0.15
    -0.02
    291
    +0.14
    0.08
    Negative Logits
    à³
    -2.08
    à²
    -1.97
    à©
    -1.72
    Õ¸
    -1.72
    à¨
    -1.64
    Õ¥
    -1.60
    Õ¡
    -1.57
    áĢº
    -1.56
    àª
    -1.56
    Õ¡Õ
    -1.51
    POSITIVE LOGITS
    oretic
    2.12
     alone
    1.80
     advantage
    1.62
    entric
    1.59
     plus
    1.53
    ~=
    1.52
    imum
    1.51
    googleapis
    1.51
     dependent
    1.50
    ographically
    1.42
    Activations Density 2.647%

    No Known Activations