INDEX

Explanations

phrases related to self-defense

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

 XIII

-0.82

 Ashe

-0.81

 GOODMAN

-0.72

ICAN

-0.71

 Nights

-0.70

 Seasons

-0.66

IUM

-0.66

 Orchestra

-0.65

BST

-0.65

mingham

-0.64

POSITIVE LOGITS

destruct

1.32

same

1.06

contained

1.02

conscious

0.99

 destruct

0.98

absor

0.95

proclaimed

0.94

awareness

0.94

esteem

0.93

 esteem

0.93

Activations Density 1.535%

phrases related to self-defense

No Comments

No Known Activations