© 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. 14682
    Prev
    Next
    INDEX
    Explanations

    phrases that attribute information to its source or reference

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "citing" related to reports or sources of information

    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
    talk
    -0.73
    ould
    -0.72
    igmatic
    -0.69
    à¦
    -0.66
    carry
    -0.66
    ascript
    -0.65
    visible
    -0.65
    oot
    -0.65
    trap
    -0.63
    spir
    -0.62
    POSITIVE LOGITS
     citing
    0.86
     unnamed
    0.77
     sources
    0.75
    uana
    0.74
     warnings
    0.74
     exemptions
    0.72
     quoting
    0.70
     exceptions
    0.68
     inconsistencies
    0.63
     unspecified
    0.62
    Activations Density 0.024%

    No Known Activations