INDEX

Explanations

words related to opinions or strong emotions about various topics

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

Elsewhere

-0.72

DIT

-0.68

emi

-0.67

dm

-0.66

cellaneous

-0.66

eway

-0.65

 ðŁĺ

-0.65

dp

-0.63

PI

-0.63

inki

-0.63

POSITIVE LOGITS

 dogged

0.72

 bipartisan

0.71

 fascination

0.69

 innocence

0.67

 fascinated

0.66

 dogma

0.64

 colonialism

0.64

itutional

0.64

 unsuccessfully

0.64

 corrid

0.64

Activations Density 0.413%

words related to opinions or strong emotions about various topics

No Comments

No Known Activations