© 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
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 28573
    Prev
    Next
    INDEX
    Explanations

    instances of coding and programming-related topics

    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
    369
    +0.16
    0.9%
    152
    +0.11
    0.6%
    95
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    152
    +0.16
    0.02
    366
    +0.11
    0.02
    195
    +0.11
    0.01
    Negative Logits
    pts
    -1.66
    ¥
    -1.58
     entry
    -1.50
     quarters
    -1.45
    ľ
    -1.45
    ]{.
    -1.44
     lux
    -1.43
    yz
    -1.42
     entries
    -1.38
    hti
    -1.35
    POSITIVE LOGITS
     argu
    1.90
    agne
    1.52
    ando
    1.47
     yours
    1.44
     (@
    1.44
    olis
    1.43
     builder
    1.42
     foreach
    1.41
    akov
    1.37
    >>>
    1.36
    Activations Density 0.036%

    No Known Activations