INDEX

Explanations

statistics and information related to crimes and legal issues, particularly those concerning violence against women

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

 Consortium

-0.73

 Unreal

-0.66

 Nost

-0.64

 Fine

-0.62

 Associated

-0.62

 Torn

-0.60

 Anonymous

-0.60

 Resistance

-0.59

 Binding

-0.59

 understatement

-0.58

POSITIVE LOGITS

've

1.07

 encount

1.06

 undertook

0.99

'd

0.98

're

0.97

 knew

0.93

 embarked

0.91

 transitioned

0.90

 entered

0.89

 exited

0.88

Activations Density 0.952%

statistics and information related to crimes and legal issues, particularly those concerning violence against women

No Comments

No Known Activations