© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    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. 6466
    Prev
    Next
    INDEX
    Explanations

    words related to subcategories, subcomponents, and subversions

    oai_token-act-pair · gpt-3.5-turbo

    references to subsystems and their characteristics

    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
    hens
    -0.78
    archment
    -0.78
    prising
    -0.76
    PIN
    -0.71
    days
    -0.71
     Colomb
    -0.70
    ello
    -0.70
     Beans
    -0.69
    atonin
    -0.69
     Flowers
    -0.68
    POSITIVE LOGITS
     sub
    3.49
     subs
    2.27
     Sub
    2.09
    Sub
    2.06
    sub
    1.89
     subp
    1.70
     SUB
    1.69
     subt
    1.42
     subst
    1.41
     subdiv
    1.38
    Activations Density 0.009%

    No Known Activations