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. 0-RES-JB
    6. 22118
    Prev
    Next
    INDEX
    Explanations

    the word "cha" or similar variations in different contexts. Since the specific word doesn't provide clear information, we can't accurately summarize without more context

    oai_token-act-pair · gpt-3.5-turbo

    the word "cha" in various contexts, indicating a focus on its occurrences within different phrases or scenarios

    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
     relocation
    -0.65
     GDDR
    -0.63
     threaded
    -0.59
    EST
    -0.59
     pedest
    -0.58
     Shore
    -0.58
     Homes
    -0.58
     retail
    -0.58
     shore
    -0.57
     Images
    -0.57
    POSITIVE LOGITS
    cha
    5.03
    CHA
    1.80
    chu
    1.77
    chin
    1.43
    chi
    1.39
    ch
    1.35
    cham
    1.27
    acha
    1.26
    che
    1.25
    chal
    1.23
    Activations Density 0.011%

    No Known Activations