INDEX

Explanations

references to historical events and political figures or discussions

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

ãĥ³ãĤ¸

-0.86

ãĤ¤ãĥĪ

-0.77

Reply

-0.65

LinkedIn

-0.63

detail

-0.62

IBLE

-0.62

Catalog

-0.61

SHARE

-0.60

 upside

-0.59

 behalf

-0.59

POSITIVE LOGITS

 arrival

0.94

 arrived

0.92

 arrives

0.89

 completes

0.86

 inception

0.84

gate

0.84

 announcement

0.83

 fiasco

0.83

 takeover

0.83

 inauguration

0.83

Activations Density 0.423%

references to historical events and political figures or discussions

No Comments

No Known Activations