INDEX

Explanations

phrases related to shifting or maintaining a focus

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

ston

-0.71

OUGH

-0.69

 paradise

-0.68

Haunted

-0.66

apons

-0.64

 Emirates

-0.61

lyn

-0.61

mia

-0.60

 attest

-0.60

 Casino

-0.60

POSITIVE LOGITS

 focus

0.94

peed

0.89

 attention

0.89

 focuses

0.87

 Attention

0.86

 focused

0.86

focus

0.84

rite

0.82

 squarely

0.81

phasis

0.78

Activations Density 2.167%

phrases related to shifting or maintaining a focus

No Comments

No Known Activations