INDEX

Explanations

words related to political figures

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

liness

-0.98

ly

-0.94

LY

-0.91

manship

-0.91

lies

-0.85

lers

-0.79

lessly

-0.78

lines

-0.77

leness

-0.76

land

-0.75

POSITIVE LOGITS

uthor

1.14

ibaba

1.02

emia

0.95

qua

0.84

 Pradesh

0.83

 Devi

0.83

isa

0.82

plings

0.81

 Haram

0.81

emon

0.81

Activations Density 0.023%

words related to political figures

No Comments

No Known Activations