INDEX

Explanations

adjectives or nouns related to visual or spatial descriptions

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

 Disorder

-0.67

Ranked

-0.66

\\\\\\\\

-0.65

akia

-0.64

ECA

-0.64

terday

-0.64

rosso

-0.62

iolet

-0.62

edIn

-0.61

 Marketable

-0.60

POSITIVE LOGITS

iest

1.28

 itself

1.22

liest

1.19

osphere

1.18

ultimate

1.01

 portion

0.92

 closest

0.86

hest

0.86

 nearest

0.83

 aspect

0.80

Activations Density 2.425%

adjectives or nouns related to visual or spatial descriptions

No Comments

No Known Activations