INDEX

Explanations

items or items related to replicas or imitations

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

ional

-0.79

RED

-0.74

ivities

-0.69

ational

-0.66

rano

-0.66

ranging

-0.66

emale

-0.65

uture

-0.65

heid

-0.64

 Scotia

-0.63

POSITIVE LOGITS

down

1.16

 knock

0.97

 down

0.87

downs

0.84

DOWN

0.81

 unconscious

0.78

ety

0.78

ı

0.78

kn

0.76

out

0.76

Activations Density 0.019%

items or items related to replicas or imitations

No Comments

No Known Activations