INDEX

Explanations

words related to instructions or actions

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

biz

-0.73

MpServer

-0.71

 SHARES

-0.69

conom

-0.69

uku

-0.68

PRESS

-0.66

BUG

-0.66

isan

-0.65

ealous

-0.65

uren

-0.65

POSITIVE LOGITS

 circulation

0.94

 shelves

0.80

pload

0.80

 fingertips

0.79

 landfill

0.76

 fireplace

0.74

 forehead

0.74

 bloodstream

0.71

 disposal

0.70

 shoulders

0.70

Activations Density 3.041%

words related to instructions or actions

No Comments

No Known Activations