© 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. 4-RES-JB
    11. 12560
    Prev
    Next
    INDEX
    Explanations

    terms related to accidents, fatalities, and criminal/legal issues

    oai_token-act-pair · gpt-3.5-turbo

    terms associated with legal or penal contexts

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

    No Comments

    Negative Logits
    thora
    -0.98
    £ı
    -0.92
    eele
    -0.84
    ajo
    -0.79
    okin
    -0.78
    tera
    -0.78
    aida
    -0.77
    sonian
    -0.75
    heast
    -0.74
    oS
    -0.72
    POSITIVE LOGITS
    ously
    0.89
     convict
    0.82
     manslaughter
    0.81
     inflicted
    0.80
     homicide
    0.77
     punishable
    0.75
    soever
    0.74
     prisoner
    0.73
     confessions
    0.72
     prisoners
    0.72
    Activations Density 0.130%

    No Known Activations