INDEX

Explanations

information related to political and government interactions and decisions

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

 tornado

-0.84

 Shelby

-0.83

 sucker

-0.82

 darn

-0.81

 raft

-0.80

 elect

-0.79

 dime

-0.79

 Everett

-0.79

 whiskey

-0.78

 Vanderbilt

-0.78

POSITIVE LOGITS

However

1.73

According

1.72

Refer

1.67

Meanwhile

1.65

Moreover

1.65

Furthermore

1.61

Regarding

1.57

Following

1.55

Besides

1.54

Apart

1.52

Activations Density 0.322%

information related to political and government interactions and decisions

No Comments

No Known Activations