INDEX

Explanations

information related to law enforcement and security preparations for events or situations

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

itol

-0.66

uta

-0.62

ocl

-0.60

¬¼

-0.59

itialized

-0.59

thia

-0.59

²¾

-0.56

 Zeal

-0.56

ãģł

-0.56

ulla

-0.56

POSITIVE LOGITS

 according

1.49

according

1.15

 says

1.06

 said

0.98

 writes

0.94

 respectively

0.91

 analysts

0.88

 experts

0.87

 prompting

0.85

said

0.83

Activations Density 1.779%

information related to law enforcement and security preparations for events or situations

No Comments

No Known Activations