INDEX

Explanations

information related to conflict, human rights abuses, and mistreatment documented in reports

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

caster

-0.73

username

-0.72

çĶ°

-0.70

linger

-0.67

Android

-0.66

 attm

-0.66

 ingred

-0.66

lly

-0.66

invoke

-0.66

(@

-0.65

POSITIVE LOGITS

 accordance

1.32

 captivity

1.18

 lieu

1.13

 spite

1.09

humane

1.06

 relation

0.97

animate

0.97

 shelters

0.97

 situ

0.95

 vitro

0.95

Activations Density 0.290%

information related to conflict, human rights abuses, and mistreatment documented in reports

No Comments

No Known Activations