INDEX

Explanations

phrases related to emergency preparedness, medical conditions, and potential conflicts like nuclear war

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĥĥãĥĪ

-0.82

naissance

-0.71

hof

-0.69

ĸļ

-0.68

ĪĴ

-0.68

 STATES

-0.66

İĭ

-0.66

 behold

-0.65

 separatist

-0.64

 Incarnation

-0.64

POSITIVE LOGITS

atial

1.23

aghetti

1.18

ooky

1.10

rawl

1.10

rit

1.09

encer

1.01

ruce

1.01

rocket

0.98

anish

0.98

itation

0.96

Activations Density 1.654%

phrases related to emergency preparedness, medical conditions, and potential conflicts like nuclear war

No Comments

No Known Activations