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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & 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. 9-RES-JB
    6. 18743
    Prev
    Next
    INDEX
    Explanations

    legal vocabulary related to criminal trials and court proceedings

    oai_token-act-pair · gpt-3.5-turbo

    legal terminology related to murder trials and convictions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 9-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    9
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    natureconservancy
    -0.90
    ebus
    -0.77
     cov
    -0.72
    achu
    -0.70
    reddits
    -0.67
     Fedora
    -0.66
     coaster
    -0.66
     Ivanka
    -0.66
    edia
    -0.65
     peppers
    -0.64
    POSITIVE LOGITS
     tribunal
    1.55
     jury
    1.54
     court
    1.54
     Court
    1.50
    Court
    1.41
     Tribunal
    1.40
     trial
    1.31
     Jury
    1.26
     verdict
    1.26
    Judge
    1.25
    Activations Density 0.484%

    No Known Activations