INDEX

Explanations

words related to legal cases and investigations, including terms like charges, court records, and authorities

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

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

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hug

-0.80

 regener

-0.80

 embrace

-0.79

 manageable

-0.76

 sustainability

-0.76

 patience

-0.76

 happiness

-0.75

 challeng

-0.74

 pace

-0.74

 arenas

-0.74

POSITIVE LOGITS

 According

1.46

 Investigators

1.43

 Presumably

1.41

 Apparently

1.40

 Sources

1.36

Sources

1.30

 However

1.28

 Reported

1.27

 Furthermore

1.26

According

1.26

Activations Density 0.546%

words related to legal cases and investigations, including terms like charges, court records, and authorities

No Comments

No Known Activations