© 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. 8-RES-JB
    11. 15142
    Prev
    Next
    INDEX
    Explanations

    descriptions of legal investigations and criminal activities

    oai_token-act-pair · gpt-3.5-turbo

    topics related to crimes and violence

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     challeng
    -0.66
     horizont
    -0.63
     agre
    -0.62
    FTWARE
    -0.60
    eday
    -0.60
     mathemat
    -0.59
     cryst
    -0.59
    angelo
    -0.57
    entary
    -0.57
    ichever
    -0.57
    POSITIVE LOGITS
     perpetrated
    0.92
     thugs
    0.88
     rape
    0.85
     thug
    0.85
     crimes
    0.85
     pedoph
    0.84
     rapists
    0.84
    ocaust
    0.84
     torture
    0.84
     killings
    0.83
    Activations Density 1.298%

    No Known Activations