INDEX

Explanations

personal pronouns indicating interaction between individuals

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

majority

-0.64

ĸļ

-0.59

iston

-0.59

AMI

-0.58

 Pastebin

-0.57

°

-0.57

ories

-0.57

nce

-0.57

 Siege

-0.55

IRC

-0.55

POSITIVE LOGITS

atic

1.10

selves

1.08

atically

1.07

 personally

0.95

self

0.90

 selves

0.88

lees

0.85

atics

0.84

adows

0.83

soever

0.82

Activations Density 1.339%

personal pronouns indicating interaction between individuals

No Comments

No Known Activations