INDEX

Explanations

phrases related to data accessibility and sharing

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

osit

-0.76

acci

-0.74

 squad

-0.64

cham

-0.63

uda

-0.61

sis

-0.60

set

-0.60

aura

-0.59

asus

-0.59

 Investigator

-0.59

POSITIVE LOGITS

ible

1.14

ibility

1.10

ibly

0.99

isable

0.85

IBLE

0.76

accessible

0.75

atable

0.73

ABLE

0.71

itable

0.70

 accessible

0.70

Activations Density 0.041%

phrases related to data accessibility and sharing

No Comments

No Known Activations