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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 29707
    Prev
    Next
    INDEX
    Explanations

    references to data structures and their attributes

    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
    271
    +0.21
    1.2%
    23
    +0.18
    1.0%
    478
    +0.16
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    23
    +0.21
    0.17
    271
    +0.18
    -0.04
    508
    +0.16
    0.11
    Negative Logits
    stown
    -1.73
    POINT
    -1.51
     Academ
    -1.46
     Applications
    -1.46
     Films
    -1.39
    Simplify
    -1.37
     avenues
    -1.35
     Sciences
    -1.35
     Hospit
    -1.32
     prospects
    -1.32
    POSITIVE LOGITS
    inated
    1.69
    bbe
    1.58
    ãĤĪãģĨ
    1.57
    quired
    1.53
    reth
    1.49
     '$
    1.47
    aes
    1.47
    ically
    1.45
    aterally
    1.44
     lack
    1.44
    Activations Density 3.655%

    No Known Activations