INDEX

Explanations

words related to criticism or negative judgment

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ership

-0.66

vention

-0.65

manship

-0.63

phrine

-0.63

 Lans

-0.62

 vitro

-0.61

 Enforcement

-0.60

doors

-0.59

ezvous

-0.58

)=(

-0.58

POSITIVE LOGITS

umpy

1.29

udge

1.27

udging

1.16

aciously

1.12

iffin

1.12

ained

1.12

anny

1.09

iddle

1.09

ace

1.07

ubby

1.06

Activations Density 0.010%

words related to criticism or negative judgment

No Comments

No Known Activations