INDEX

Explanations

phrases related to decision-making or taking action

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_fs12288-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

12,288

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

 controvers

-0.56

 Alpine

-0.56

 municip

-0.55

 multinational

-0.54

virt

-0.53

 Infrastructure

-0.52

 Proposition

-0.52

 strikingly

-0.52

 Historically

-0.52

 laboratories

-0.51

POSITIVE LOGITS

 gonna

1.13

 fuckin

1.12

 haha

1.01

 kinda

1.00

 wanna

0.97

 fucking

0.96

laughs

0.91

ya

0.91

..."

0.91

 somet

0.90

Activations Density 1.763%

phrases related to decision-making or taking action

No Comments

No Known Activations