INDEX

Explanations

references to locations and specific details in news articles

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

Rav

-0.59

cru

-0.57

ham

-0.57

shire

-0.56

itiz

-0.56

 compan

-0.56

orship

-0.56

spr

-0.56

stru

-0.55

 shaped

-0.55

POSITIVE LOGITS

+.

1.02

+)

0.92

+,

0.84

ABV

0.84

-+

0.83

/-

0.76

ablishment

0.74

âĨĳ

0.73

/+

0.72

Activations Density 0.286%

references to locations and specific details in news articles

No Comments

No Known Activations