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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 29091
    Prev
    Next
    INDEX
    Explanations

    connections and associations within a network

    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
    260
    +0.13
    0.7%
    127
    +0.12
    0.7%
    494
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    372
    +0.13
    0.01
    260
    +0.12
    0.02
    148
    +0.12
    0.01
    Negative Logits
    akers
    -1.64
    orer
    -1.58
    endment
    -1.56
    »
    -1.55
    clamation
    -1.52
     slogan
    -1.47
    Ĩ
    -1.46
    Competing
    -1.40
    ģ
    -1.39
     excuse
    -1.34
    POSITIVE LOGITS
     thereto
    1.98
    aterally
    1.65
    ely
    1.64
    rier
    1.63
    ioned
    1.60
    ues
    1.57
    thon
    1.57
    itudinal
    1.56
     hereto
    1.55
    vest
    1.52
    Activations Density 0.145%

    No Known Activations