INDEX

Explanations

words related to body image and unrealistic standards in society, particularly within the gay community

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

ILLE

-0.84

 Zurich

-0.72

rot

-0.69

 Walker

-0.68

 Muse

-0.67

 Moody

-0.67

 GOODMAN

-0.67

 Messi

-0.66

OHN

-0.66

 Bunny

-0.64

POSITIVE LOGITS

missive

1.45

sequent

1.38

sequently

1.37

stantial

1.34

verting

1.31

mitter

1.25

ordinate

1.24

mitted

1.23

urban

1.16

Redd

1.16

Activations Density 0.314%

words related to body image and unrealistic standards in society, particularly within the gay community

No Comments

No Known Activations