INDEX

Explanations

information related to instructions or guidelines

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

otle

-0.82

gerald

-0.78

isconsin

-0.75

etz

-0.74

aukee

-0.71

amaru

-0.70

utra

-0.70

ysical

-0.69

thouse

-0.68

enegger

-0.68

POSITIVE LOGITS

nd

1.91

 thirds

1.48

 halves

1.20

ND

1.10

0.84

 externalToEVAOnly

0.84

0.83

0.81

 sides

0.80

 tablespoons

0.75

Activations Density 1.515%

information related to instructions or guidelines

No Comments

No Known Activations