INDEX

Explanations

information about physical harm or damage caused by accidents, conflicts, or attacks

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

ullivan

-0.91

 liner

-0.71

ramid

-0.69

zsche

-0.69

perm

-0.66

soDeliveryDate

-0.65

archs

-0.64

 Prospect

-0.63

SpaceEngineers

-0.62

arch

-0.62

POSITIVE LOGITS

 inflicted

1.32

 sustained

1.01

 suffered

0.94

 incurred

0.91

 injuries

0.87

 wounds

0.86

staking

0.85

bite

0.83

 healed

0.81

 injury

0.81

Activations Density 0.940%

information about physical harm or damage caused by accidents, conflicts, or attacks

No Comments

No Known Activations