INDEX

Explanations

comparative phrases expressing resemblance

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

alez

-0.77

alt

-0.77

inion

-0.77

byn

-0.72

utherland

-0.72

irtual

-0.72

arcity

-0.71

rax

-0.71

itles

-0.71

otom

-0.71

POSITIVE LOGITS

lier

1.17

liest

1.07

 crap

1.03

lihood

1.00

 shit

0.78

 filler

0.72

an

0.70

gib

0.69

 something

0.69

Activations Density 0.511%

comparative phrases expressing resemblance

No Comments

No Known Activations