INDEX

Explanations

references to politicians, political leaders, and political ideologies or comparisons

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

 grate

-0.85

Director

-0.72

trap

-0.72

Effects

-0.70

gallery

-0.67

iera

-0.67

cookie

-0.64

reci

-0.63

bum

-0.63

BI

-0.63

POSITIVE LOGITS

 Opposition

0.99

 opposition

0.96

 revolt

0.89

 Gamergate

0.89

 insurgent

0.83

 opposing

0.82

 FIGHT

0.80

 Confederation

0.79

 resistance

0.78

odox

0.77

Activations Density 0.075%

references to politicians, political leaders, and political ideologies or comparisons

No Comments

No Known Activations