INDEX

Explanations

names of individuals or locations followed by a colon

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

 overl

-0.72

ilater

-0.68

inement

-0.67

 tremend

-0.66

æ©

-0.66

 bloss

-0.65

aband

-0.64

 arous

-0.64

 troubles

-0.62

 reconnect

-0.62

POSITIVE LOGITS

 Yeah

0.96

 Alright

0.95

Why

0.86

Fact

0.85

Hey

0.84

 Would

0.84

 Calling

0.82

 What

0.82

 "...

0.82

 Okay

0.80

Activations Density 0.115%

names of individuals or locations followed by a colon

No Comments

No Known Activations

names of individuals or locations followed by a colon

No Comments

No Known Activations