INDEX

Explanations

phrases related to pathways or directions

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

asts

-0.81

eg

-0.81

aples

-0.76

ilts

-0.75

 livest

-0.71

eals

-0.70

ines

-0.66

rongh

-0.66

etheus

-0.66

 streng

-0.65

POSITIVE LOGITS

 whatsoever

0.86

point

0.77

 anymore

0.76

 anybody

0.72

 else

0.67

ouver

0.66

fare

0.66

aterasu

0.65

shake

0.65

 anyone

0.64

Activations Density 0.023%

phrases related to pathways or directions

No Comments

No Known Activations