INDEX

Explanations

personal pronouns and verbs indicating speaking, writing, or explaining from a first-person perspective

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

noon

-0.73

acters

-0.69

estial

-0.63

ictional

-0.61

rocket

-0.60

selection

-0.60

iencies

-0.60

 torch

-0.59

paralle

-0.59

 cloning

-0.58

POSITIVE LOGITS

 said

1.47

said

1.25

 wrote

1.23

 says

1.23

 exclaimed

1.21

 explained

1.14

Said

1.13

 replied

1.13

 joked

1.10

 told

1.09

Activations Density 0.836%

personal pronouns and verbs indicating speaking, writing, or explaining from a first-person perspective

No Comments

No Known Activations