INDEX

Explanations

steps or stages in a process or set of actions

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

inately

-0.80

ores

-0.79

anguages

-0.70

ecause

-0.69

iqueness

-0.68

rums

-0.67

olls

-0.67

ciating

-0.67

okia

-0.66

gdala

-0.66

POSITIVE LOGITS

hens

1.08

daughter

1.01

Step

0.98

 Steps

0.97

steps

0.95

 steps

0.93

isters

0.93

step

0.93

 step

0.81

antry

0.81

Activations Density 2.108%

steps or stages in a process or set of actions

No Comments

No Known Activations