Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 4-RES-JB
    6. 18085
    Prev
    Next
    INDEX
    Explanations

    the phrase "good news"

    oai_token-act-pair · gpt-3.5-turbo

    instances of "good news."

    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
    aval
    -0.75
    urat
    -0.72
    apons
    -0.71
    senal
    -0.71
     solvent
    -0.69
     sediment
    -0.65
    ngth
    -0.65
    entric
    -0.64
     trained
    -0.63
    rowd
    -0.63
    POSITIVE LOGITS
    worthiness
    0.91
    worthy
    0.90
    reader
    0.79
    peak
    0.74
    Catholic
    0.70
     headlines
    0.70
    night
    0.70
    tainment
    0.70
    telling
    0.69
    ween
    0.66
    Activations Density 0.029%

    No Known Activations