INDEX

Explanations

words related to the concept of "rogue"

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

birth

-0.65

 goodbye

-0.62

Fas

-0.60

AAP

-0.60

aved

-0.59

 Machina

-0.55

 spaced

-0.55

 shorth

-0.54

Fey

-0.54

 Wasserman

-0.54

POSITIVE LOGITS

raphic

1.42

raphics

1.28

raph

1.21

allery

1.10

roup

1.09

aming

1.02

ressive

0.99

rams

0.99

roups

0.98

rowth

0.97

Activations Density 0.018%

words related to the concept of "rogue"

No Comments

No Known Activations