INDEX

Explanations

phrases related to continuous actions or processes

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

arta

-0.78

ooters

-0.74

aster

-0.73

oster

-0.72

*/(

-0.70

onom

-0.69

Ã©s

-0.67

RN

-0.67

ustomed

-0.67

RNA

-0.67

POSITIVE LOGITS

 unab

0.98

 uninterrupted

0.92

 onward

0.82

 indefinitely

0.79

 progressing

0.74

 unchanged

0.74

 hostilities

0.73

 onwards

0.72

ap

0.72

 unanswered

0.69

Activations Density 0.510%

phrases related to continuous actions or processes

No Comments

No Known Activations