INDEX

Explanations

quotations

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

nown

-0.81

ensor

-0.74

HAEL

-0.72

idious

-0.67

ENS

-0.66

 Engel

-0.63

anie

-0.62

odcast

-0.62

ENSE

-0.62

ographed

-0.61

POSITIVE LOGITS

Cause

1.01

Mech

0.91

taboola

0.89

cause

0.86

tis

0.79

Interstitial

0.79

MpServer

0.78

mond

0.77

yer

0.75

til

0.74

Activations Density 3.902%

quotations

No Comments

No Known Activations