INDEX

Explanations

information related to organizational members or associates

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

enegger

-0.72

anche

-0.68

 msec

-0.67

trop

-0.64

bows

-0.60

ondo

-0.59

urses

-0.59

 Tropical

-0.58

urse

-0.58

ijn

-0.56

POSITIVE LOGITS

hips

1.50

hip

1.04

doms

1.01

member

0.96

 dues

0.94

board

0.83

members

0.80

ials

0.79

holder

0.78

ority

0.78

Activations Density 0.568%

information related to organizational members or associates

No Comments

No Known Activations