INDEX

Explanations

proper nouns and names, particularly related to movies, actors, politicians, and places

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

rane

-0.73

mble

-0.61

 Bowen

-0.59

erella

-0.59

 critically

-0.59

aspers

-0.59

Ļ½

-0.58

ollen

-0.58

ãĥ¼ãĥĨ

-0.57

 Perspect

-0.56

POSITIVE LOGITS

 whatsoever

1.35

nor

0.82

brainer

0.76

ody

0.72

onsense

0.71

*/(

0.71

hawk

0.66

 anymore

0.66

 dime

0.65

 hesitation

0.64

Activations Density 0.093%

proper nouns and names, particularly related to movies, actors, politicians, and places

No Comments

No Known Activations