INDEX

Explanations

names of people or entities

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

mble

-0.85

ascus

-0.81

ional

-0.77

istors

-0.73

ality

-0.73

okÃ©

-0.69

iment

-0.68

ience

-0.67

aciously

-0.67

acious

-0.67

POSITIVE LOGITS

mares

1.37

mare

1.30

 Templar

1.25

fighter

0.91

 Templ

0.89

 Riders

0.88

doms

0.88

 Rider

0.83

holder

0.81

fish

0.80

Activations Density 0.025%

names of people or entities

No Comments

No Known Activations