Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 11445
    Prev
    Next
    INDEX
    Explanations

    references to SQL Server and related technologies

    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
    330
    +0.12
    0.7%
    357
    +0.11
    0.6%
    72
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    163
    +0.12
    0.01
    357
    +0.11
    0.01
    13
    +0.11
    0.02
    Negative Logits
    IJ
    -3.49
    ·
    -3.49
    ³
    -3.34
    ¸
    -3.33
    ·¸
    -3.28
    į
    -3.19
    ¶
    -3.16
    ¦
    -3.06
    Ħ
    -3.05
    ľ
    -3.03
    POSITIVE LOGITS
    ERTY
    1.85
     Management
    1.52
    chemy
    1.51
     knows
    1.50
     exists
    1.49
    inux
    1.49
     existed
    1.45
     itself
    1.44
     anymore
    1.42
     Trading
    1.40
    Activations Density 0.218%

    No Known Activations