© 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. 20309
    Prev
    Next
    INDEX
    Explanations

    phrases indicating a technical issue or error messages

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to error messages and retry prompts

    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
    roots
    -0.74
    ³³³³³³³³
    -0.66
    ³³³³³³³³³³³³³³³³
    -0.64
    ³³³³
    -0.63
    etts
    -0.59
     Instruments
    -0.59
     Penal
    -0.59
    toe
    -0.57
    owship
    -0.56
    pires
    -0.56
    POSITIVE LOGITS
     Invalid
    0.99
    atta
    0.68
    ãģ®
    0.68
    ality
    0.67
    ally
    0.63
    ãĤ¨ãĥ«
    0.62
     delete
    0.62
    icol
    0.60
    姫
    0.59
    cooked
    0.59
    Activations Density 0.021%

    No Known Activations