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. 29475
    Prev
    Next
    INDEX
    Explanations

    phrases related to systemic evaluation and oversight procedures

    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
    56
    +0.14
    0.8%
    425
    +0.12
    0.7%
    199
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    425
    +0.14
    0.13
    259
    +0.12
    0.03
    199
    +0.11
    0.04
    Negative Logits
    actin
    -1.59
    áĢº
    -1.56
    á̝
    -1.48
    odot
    -1.42
    =(
    -1.40
    ophagus
    -1.39
    odium
    -1.37
    UTERS
    -1.33
    astom
    -1.32
    haps
    -1.32
    POSITIVE LOGITS
     corner
    1.55
    ©
    1.54
    ties
    1.45
    §
    1.43
     reinst
    1.41
    °
    1.39
    ļ
    1.37
    ĺ
    1.35
     [@
    1.33
    ĸ
    1.32
    Activations Density 3.924%

    No Known Activations