INDEX

Explanations

phrases related to challenges or difficulties

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

allery

-0.72

ilater

-0.71

ividual

-0.71

amera

-0.70

 Loft

-0.69

umbn

-0.69

oir

-0.66

aurus

-0.64

alian

-0.64

Tot

-0.62

POSITIVE LOGITS

ening

1.10

ened

1.05

coded

0.99

wired

0.93

nesses

0.91

ball

0.89

cover

0.89

core

0.86

 enough

0.85

eners

0.85

Activations Density 3.849%

phrases related to challenges or difficulties

No Comments

No Known Activations