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. 29262
    Prev
    Next
    INDEX
    Explanations

    terms related to contamination and environmental standards

    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
    203
    +0.18
    1.0%
    285
    +0.14
    0.8%
    164
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    203
    +0.18
    0.03
    164
    +0.14
    0.03
    285
    +0.14
    0.02
    Negative Logits
    "));
    -1.75
    yer
    -1.66
    ymoon
    -1.62
    "),
    -1.52
    hower
    -1.51
    ward
    -1.48
    )$)
    -1.47
    )$-
    -1.45
    ÅĤa
    -1.41
    ")
    -1.40
    POSITIVE LOGITS
    ortion
    1.66
     reached
    1.58
    raction
    1.54
    asting
    1.51
    oon
    1.43
     tilt
    1.41
    ESE
    1.41
    ain
    1.39
     view
    1.36
    iev
    1.35
    Activations Density 0.130%

    No Known Activations