© 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. 3-RES-JB
    11. 15117
    Prev
    Next
    INDEX
    Explanations

    references to legal terms and courtroom proceedings

    oai_token-act-pair · gpt-3.5-turbo

    references to jury decisions and actions in legal contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    irst
    -0.72
     Franch
    -0.70
     Occupations
    -0.69
     Prosper
    -0.64
     Horus
    -0.63
    eworld
    -0.60
    dayName
    -0.60
    kun
    -0.59
     tradem
    -0.59
    erenn
    -0.58
    POSITIVE LOGITS
     jury
    1.02
     Jury
    0.89
     jurors
    0.88
     verdict
    0.83
    glass
    0.80
    proof
    0.78
    INAL
    0.75
    naire
    0.74
    Panel
    0.73
     commissioners
    0.71
    Activations Density 0.011%

    No Known Activations