INDEX

Explanations

phrases related to escalating conflicts or tensions

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

itte

-0.89

ieu

-0.84

lean

-0.82

leans

-0.80

girl

-0.80

inates

-0.79

leck

-0.78

print

-0.77

prints

-0.77

ndra

-0.74

POSITIVE LOGITS

 tensions

1.20

 hostilities

1.16

 escalate

1.05

 escal

1.02

 escalation

0.98

 tension

0.95

 escalating

0.94

 confrontation

0.93

 frustration

0.87

 frustrations

0.86

Activations Density 0.055%

phrases related to escalating conflicts or tensions

No Comments

No Known Activations