INDEX

Explanations

steps or instructions for performing a technical task, possibly related to data or file management

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

wives

-0.85

IVERS

-0.76

 virginity

-0.73

von

-0.71

giving

-0.70

wife

-0.70

isky

-0.66

conservancy

-0.65

 affairs

-0.62

venge

-0.62

POSITIVE LOGITS

 button

1.29

 buttons

1.23

 toolbar

1.16

button

1.15

 scroll

1.13

 cursor

1.10

 icon

1.09

 popup

1.07

 scrolling

1.02

 arrow

1.02

Activations Density 2.226%

steps or instructions for performing a technical task, possibly related to data or file management

No Comments

No Known Activations