INDEX

Explanations

verbs describing actions and intentions

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

©¶æ¥µ

-0.85

é¾įå

-0.82

çīĪ

-0.73

 professionally

-0.72

mberg

-0.72

ETF

-0.69

avid

-0.66

NCT

-0.65

odox

-0.65

 Universities

-0.63

POSITIVE LOGITS

 itself

1.10

ITS

0.89

its

0.85

Temperature

0.75

 tricks

0.73

 electrons

0.73

 mating

0.72

 vibrations

0.71

 orbits

0.71

 humanoid

0.69

Activations Density 0.624%

verbs describing actions and intentions

No Comments

No Known Activations