INDEX

Explanations

information related to people's actions and events such as deaths, arrests, and followings

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

 Locations

-0.81

 divisions

-0.76

 deployments

-0.69

 reservoirs

-0.67

 expansions

-0.67

 shipments

-0.67

isSpecialOrderable

-0.66

 closures

-0.64

 highways

-0.64

 Regulations

-0.64

POSITIVE LOGITS

 whom

1.32

who

1.06

 himself

0.89

who

0.89

whose

0.89

 assassinated

0.84

hunt

0.80

 died

0.79

nikov

0.79

opus

0.77

Activations Density 2.858%

information related to people's actions and events such as deaths, arrests, and followings

No Comments

No Known Activations