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. 11-RES-JB
    6. 15111
    Prev
    Next
    INDEX
    Explanations

    phrases related to legal matters, especially around the exchange or withholding of information and legal proceedings

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to legal proceedings and judicial decisions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Uni
    -0.85
     Eco
    -0.81
    reddits
    -0.76
    paces
    -0.73
     Entreprene
    -0.73
    Grad
    -0.72
    romeda
    -0.72
     Forums
    -0.71
    ghai
    -0.71
    umbai
    -0.70
    POSITIVE LOGITS
     jurors
    1.75
     testimony
    1.63
     witnesses
    1.59
     testified
    1.54
     perjury
    1.51
     circumst
    1.50
     testifying
    1.48
     defendant
    1.47
     incrim
    1.44
     witness
    1.44
    Activations Density 0.691%

    No Known Activations