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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 0-RES-JB
    11. 23246
    Prev
    Next
    INDEX
    Explanations

    words related to legal documents and procedures

    oai_token-act-pair · gpt-3.5-turbo

    references to attentiveness or focus

    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
    anyon
    -0.68
    ools
    -0.65
    elist
    -0.65
    alysis
    -0.63
    ryce
    -0.63
    rored
    -0.62
    unning
    -0.62
    fect
    -0.61
    anche
    -0.61
    RY
    -0.60
    POSITIVE LOGITS
    ļéĨĴ
    1.10
     ATT
    0.98
    ãĤ¨ãĥ«
    0.88
    åij
    0.74
    ãĥł
    0.70
     ABE
    0.70
    Magikarp
    0.70
     Waterloo
    0.69
    é¾į
    0.67
    ihad
    0.66
    Activations Density 0.001%

    No Known Activations