INDEX

Explanations

phrases or keywords related to taking turns

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

Negative Logits

inately

-0.71

ecause

-0.69

hammad

-0.65

capacity

-0.63

 Lumpur

-0.63

anyahu

-0.62

ropolitan

-0.62

oreal

-0.58

llah

-0.58

 Sith

-0.58

POSITIVE LOGITS

buck

1.06

about

0.94

abouts

0.91

coat

0.84

overs

0.78

over

0.77

ips

0.71

igraph

0.71

 tides

0.71

laus

0.71

Activations Density 0.019%

phrases or keywords related to taking turns

No Known Activations