INDEX

Explanations

costumes or costume-related terms

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

venants

-0.81

utherford

-0.79

nesota

-0.73

esm

-0.72

mediate

-0.69

raltar

-0.67

itude

-0.67

nir

-0.67

orio

-0.67

POSITIVE LOGITS

 worn

0.96

 costumes

0.95

 costume

0.87

 jewelry

0.81

 wearer

0.81

 attire

0.79

 apparel

0.76

 malfunction

0.74

 wardrobe

0.73

 designer

0.73

Activations Density 0.042%

costumes or costume-related terms

No Comments

No Known Activations