INDEX

Explanations

negative situations or outcomes

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

rouse

-0.80

ynthesis

-0.77

glas

-0.75

cript

-0.72

cellent

-0.71

 Collider

-0.71

uador

-0.71

leans

-0.70

amaru

-0.70

ools

-0.69

POSITIVE LOGITS

 inflicted

0.89

 plag

0.88

 havoc

0.86

fully

0.77

 consequences

0.77

 horribly

0.77

imaru

0.76

 nightmares

0.76

der

0.74

 consequence

0.73

Activations Density 2.512%

negative situations or outcomes

No Comments

No Known Activations