INDEX

Explanations

references to technical specifications and detailed information within text, possibly related to research articles or reports

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

 anniversary

-0.70

jah

-0.69

lling

-0.68

ilion

-0.67

clus

-0.66

ework

-0.66

legged

-0.65

hemer

-0.65

archment

-0.65

ãĥĲ

-0.64

POSITIVE LOGITS

 Anyway

1.37

 Regardless

1.09

 Similarly

1.05

 Likewise

1.05

 Accordingly

1.02

 Moreover

1.02

 Nevertheless

1.01

 Instead

1.01

 Additionally

1.01

 Therefore

1.01

Activations Density 0.044%

references to technical specifications and detailed information within text, possibly related to research articles or reports

No Comments

No Known Activations