INDEX

Explanations

phrases related to the future

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

edin

-0.75

lee

-0.74

ema

-0.73

Fat

-0.72

 anecd

-0.72

 ding

-0.71

ideshow

-0.71

cano

-0.70

idelines

-0.70

zip

-0.70

POSITIVE LOGITS

 generations

1.04

izons

0.97

 prospects

0.94

 horizon

0.93

0.87

 future

0.87

 scenarios

0.86

 outlook

0.83

future

0.80

 possibilities

0.80

Activations Density 1.560%

phrases related to the future

No Comments

No Known Activations