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. 11765
    Prev
    Next
    INDEX
    Explanations

    keywords related to ethical considerations in video evidence and access regulations

    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
    17
    +0.21
    1.2%
    233
    +0.15
    0.9%
    507
    +0.15
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    111
    +0.21
    0.24
    156
    +0.15
    0.25
    71
    +0.15
    0.25
    Negative Logits
    ici
    -1.81
    erty
    -1.70
    orously
    -1.59
    ples
    -1.53
    ically
    -1.53
    otyped
    -1.53
    entially
    -1.44
    assed
    -1.44
     wise
    -1.41
    ameter
    -1.41
    POSITIVE LOGITS
    ĻĤ
    2.38
    ¼
    2.32
    ¿½
    2.27
    Ĥ
    2.16
    ł
    1.96
    »
    1.93
    ĸ´
    1.91
    ¸
    1.87
    Ĵ
    1.87
    ¹
    1.84
    Activations Density 4.472%

    No Known Activations