© 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. 11-RES-JB
    11. 826
    Prev
    Next
    INDEX
    Explanations

    information related to criminal activities and legal proceedings

    oai_token-act-pair · gpt-3.5-turbo

    phrases that indicate legal charges or convictions

    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
    unny
    -0.77
    Vo
    -0.74
    ires
    -0.73
    ees
    -0.71
    Cath
    -0.70
    wake
    -0.70
    itar
    -0.69
    å¹
    -0.68
    ascript
    -0.68
    move
    -0.67
    POSITIVE LOGITS
     manslaughter
    1.39
     murdering
    1.38
     raping
    1.34
     conspiring
    1.21
     assaulting
    1.19
     robbing
    1.12
     killing
    1.12
     felony
    1.11
     murder
    1.11
     stabbing
    1.08
    Activations Density 0.097%

    No Known Activations