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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 17301
    Prev
    Next
    INDEX
    Explanations

    phrases mentioning or referring to the concept of 'bits'

    oai_token-act-pair · gpt-3.5-turbo

    references to "bits" or "bit" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    naire
    -0.70
    intend
    -0.69
    aver
    -0.66
    PRO
    -0.66
    ita
    -0.65
    emi
    -0.65
     Petition
    -0.64
    ador
    -0.64
    agogue
    -0.63
    ario
    -0.63
    POSITIVE LOGITS
     bits
    3.75
     Bits
    2.36
    bits
    2.13
     chunks
    1.85
     pieces
    1.84
     bit
    1.60
     snippets
    1.54
     Pieces
    1.53
     scraps
    1.51
     fragments
    1.49
    Activations Density 0.009%

    No Known Activations