INDEX

Explanations

References to historical events or conflicts, particularly focusing on wars

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 confir

-0.82

obook

-0.74

Asset

-0.72

 mathemat

-0.67

sembly

-0.67

leased

-0.66

ucha

-0.65

erved

-0.65

prints

-0.65

ãĥ¤

-0.64

POSITIVE LOGITS

lords

1.17

lord

1.05

rior

0.97

riors

0.96

era

0.92

fare

0.89

Era

0.87

 waged

0.86

bler

0.85

 raging

0.82

Activations Density 0.033%

References to historical events or conflicts, particularly focusing on wars

No Comments

No Known Activations