INDEX

Explanations

mentions of specific geographical locations, events, and individuals related to a particular region

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ngth

-0.71

DonaldTrump

-0.70

ilial

-0.69

 Schn

-0.68

 NETWORK

-0.68

ensional

-0.61

 guid

-0.61

Magikarp

-0.61

awaru

-0.61

 oblig

-0.61

POSITIVE LOGITS

ished

1.05

aby

1.05

er

0.99

lee

0.97

burn

0.93

ards

0.92

ishing

0.91

lett

0.89

ters

0.89

iation

0.84

Activations Density 0.018%

mentions of specific geographical locations, events, and individuals related to a particular region

No Comments

No Known Activations