INDEX

Explanations

text related to reviewing, examining, assessing, and considering different policies and regulations

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

Torrent

-0.77

arak

-0.76

Cry

-0.68

cil

-0.66

soever

-0.64

guard

-0.64

WARNING

-0.63

ottest

-0.63

ARM

-0.63

init

-0.63

POSITIVE LOGITS

 whether

1.16

 feasibility

1.10

 trends

0.97

how

0.94

 possibilities

0.90

 aspects

0.89

 alternatives

0.87

 factors

0.85

whether

0.85

 behalf

0.84

Activations Density 2.674%

text related to reviewing, examining, assessing, and considering different policies and regulations

No Comments

No Known Activations