© 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. 4978
    Prev
    Next
    INDEX
    Explanations

    words related to physical actions or movements that involve force or impact

    oai_token-act-pair · gpt-3.5-turbo

    instances of the letter "h" in various contexts

    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
    ãĤ´ãĥ³
    -0.78
    terday
    -0.71
    lihood
    -0.69
     flush
    -0.65
     mine
    -0.64
    ãĥ¯
    -0.63
    Ô
    -0.62
    Leaks
    -0.61
     Beware
    -0.61
     conclud
    -0.60
    POSITIVE LOGITS
    oused
    1.27
    anky
    1.18
    anging
    1.18
    ulk
    1.16
    ousing
    1.13
    uddled
    1.13
    ospital
    1.12
    ollywood
    1.11
    acking
    1.10
    olly
    1.10
    Activations Density 0.013%

    No Known Activations