INDEX

Explanations

phrases related to political intrigue and conspiracy

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

 parentheses

-0.69

paren

-0.67

 Quart

-0.66

 caveats

-0.65

 pauses

-0.64

noon

-0.63

 averages

-0.63

Typ

-0.62

 strides

-0.61

 anecd

-0.61

POSITIVE LOGITS

 revenge

1.08

 blackmail

1.07

 extortion

1.00

 ransom

0.96

ihadi

0.93

venge

0.93

 jihad

0.91

 nefarious

0.90

 terrorist

0.88

 assassinate

0.87

Activations Density 0.554%

phrases related to political intrigue and conspiracy

No Comments

No Known Activations