INDEX

Explanations

phrases related to negotiations and communication

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

externalActionCode

-0.71

SPONSORED

-0.68

VIEW

-0.65

livious

-0.62

 confines

-0.62

PC

-0.60

PM

-0.60

hack

-0.59

 Desc

-0.59

 experiment

-0.59

POSITIVE LOGITS

rid

0.97

tin

0.95

 plenty

0.92

cha

0.87

 Talent

0.86

ngth

0.77

 precedence

0.76

 Plenty

0.72

rontal

0.70

 tonnes

0.69

Activations Density 0.025%

phrases related to negotiations and communication

No Comments

No Known Activations