INDEX

Explanations

words related to visual perception or observation

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

ten

-0.67

ivari

-0.67

 Violent

-0.63

urdue

-0.63

 ACPI

-0.61

minus

-0.61

yan

-0.59

TON

-0.58

 Mercer

-0.58

icum

-0.57

POSITIVE LOGITS

ight

1.13

seeing

1.08

 sockets

1.04

eyed

1.03

eye

0.98

ewitness

0.95

 gaze

0.95

beam

0.93

 lenses

0.91

Eye

0.91

Activations Density 0.767%

words related to visual perception or observation

No Comments

No Known Activations