INDEX

Explanations

instances of the word "to."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

blocks

-0.69

 blocks

-0.66

 Hours

-0.66

rising

-0.65

updated

-0.61

 Closing

-0.60

ivated

-0.59

kind

-0.59

hered

-0.59

rio

-0.58

POSITIVE LOGITS

 participate

1.29

 testify

1.16

 withstand

1.16

 undertake

1.11

 handle

1.08

 receive

1.06

 qualify

1.06

 cope

1.06

 partake

1.04

 engage

1.03

Activations Density 0.142%

instances of the word "to."

No Comments

No Known Activations