© 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. 0-RES-JB
    11. 11627
    Prev
    Next
    INDEX
    Explanations

    phrases related to protective or supportive structures or surfaces, such as 'maternal dynasty'

    oai_token-act-pair · gpt-3.5-turbo

    references to "mat" or "mats," possibly indicating a focus on certain physical objects or contexts where they are used

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    hire
    -0.86
    HF
    -0.69
    eeks
    -0.67
    burgh
    -0.65
    oa
    -0.65
    DD
    -0.65
     Olsen
    -0.64
     WD
    -0.62
     Shutdown
    -0.61
    iration
    -0.61
    POSITIVE LOGITS
     mat
    4.00
     mats
    2.47
     Mat
    1.90
    Mat
    1.82
    mat
    1.77
     matrix
    1.43
     MAT
    1.19
     Mats
    1.14
     matt
    1.09
    MAT
    1.06
    Activations Density 0.008%

    No Known Activations