INDEX

Explanations

time-related events or transitions in a narrative

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

aez

-0.73

orde

-0.67

onomy

-0.65

scription

-0.63

orum

-0.62

TY

-0.62

cos

-0.62

holder

-0.60

md

-0.60

own

-0.60

POSITIVE LOGITS

 proceeded

1.12

 proceed

0.86

 promptly

0.82

 succumb

0.81

 withdrew

0.80

 succumbed

0.76

 switched

0.74

 recons

0.73

 retracted

0.73

 reverted

0.72

Activations Density 0.054%

time-related events or transitions in a narrative

No Comments

No Known Activations