INDEX

Explanations

phrases related to positive outcomes or optimisms

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

eters

-0.79

oths

-0.74

opers

-0.72

hyde

-0.69

ategory

-0.69

eds

-0.68

 Pavilion

-0.67

pper

-0.66

ĸļ

-0.65

atum

-0.64

POSITIVE LOGITS

enough

1.30

reads

1.14

 luck

1.13

luck

1.11

 Samar

1.09

bye

1.09

ol

1.02

 intentions

1.01

 enough

0.92

 deed

0.91

Activations Density 0.379%

phrases related to positive outcomes or optimisms

No Comments

No Known Activations