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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 1-RES-JB
    6. 10555
    Prev
    Next
    INDEX
    Explanations

    words related to legal proceedings or actions, specifically the term "prosecute."

    oai_token-act-pair · gpt-3.5-turbo

    terms related to legal actions, specifically prosecution

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     preference
    -0.78
    lighting
    -0.76
     hypers
    -0.72
    opian
    -0.72
    lights
    -0.68
     capacitor
    -0.64
     immersion
    -0.64
    arat
    -0.62
     flyers
    -0.61
    yip
    -0.61
    POSITIVE LOGITS
     prosecuted
    1.06
     prosecute
    0.99
     Prosecut
    0.89
    heet
    0.84
     prosecut
    0.80
     prosecuting
    0.79
    Prosecut
    0.76
    iary
    0.72
    ices
    0.72
    20439
    0.69
    Activations Density 0.009%

    No Known Activations