© 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. 3949
    Prev
    Next
    INDEX
    Explanations

    descriptions of objects and their attachments

    oai_token-act-pair · gpt-3.5-turbo

    components and connections in mechanical or electronic assemblies

    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
     Samar
    -0.85
     bloodshed
    -0.76
    violence
    -0.72
    hya
    -0.69
    oké
    -0.68
     Genocide
    -0.68
     Pension
    -0.67
     Archdemon
    -0.67
    wage
    -0.67
     Violence
    -0.66
    POSITIVE LOGITS
     removable
    1.18
     protr
    1.11
     diameter
    1.07
     flats
    1.03
     molded
    1.01
    Accessory
    0.97
     angled
    0.96
     adjustable
    0.96
     securely
    0.96
     attaching
    0.95
    Activations Density 0.550%

    No Known Activations