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. 11534
    Prev
    Next
    INDEX
    Explanations

    data types and structures commonly used in programming, particularly in relation to arrays and numerical values

    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
    419
    +0.15
    0.8%
    427
    +0.15
    0.8%
    159
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    0
    +0.15
    0.13
    427
    +0.15
    0.10
    216
    +0.12
    0.16
    Negative Logits
    etter
    -1.53
    umi
    -1.50
    onomy
    -1.48
    "}](#
    -1.46
    etting
    -1.43
    sa
    -1.43
    ighters
    -1.42
    avor
    -1.41
    ild
    -1.40
    ensit
    -1.38
    POSITIVE LOGITS
    Ŀ
    2.70
    µ
    2.63
    ²
    2.59
    ¸
    2.58
    Ī
    2.56
    ¿
    2.54
    Į
    2.53
    ķ
    2.51
    ģ
    2.50
    ĥ½
    2.50
    Activations Density 5.819%

    No Known Activations