INDEX

Explanations

social media-related terms and interactions

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

ndra

-0.89

aternity

-0.84

verning

-0.72

deals

-0.70

uating

-0.67

alty

-0.67

ablished

-0.66

 hosp

-0.65

ocrats

-0.65

ente

-0.64

POSITIVE LOGITS

 footage

1.55

 recordings

1.42

 video

1.41

 videos

1.39

 uploaded

1.38

 videot

1.31

 excerpts

1.30

 screenshots

1.28

 audio

1.28

PDF

1.28

Activations Density 3.580%

social media-related terms and interactions

No Comments

No Known Activations