INDEX

Explanations

phrases detailing incidents or crimes

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 fantasy

-0.69

enance

-0.65

gey

-0.63

 intrinsically

-0.62

Imagine

-0.61

billion

-0.60

 perenn

-0.60

die

-0.60

paren

-0.60

pires

-0.59

POSITIVE LOGITS

Officers

1.08

 Officers

0.96

Police

0.93

 alerted

0.92

nsics

0.91

police

0.90

 officers

0.90

 escorted

0.90

 RCMP

0.89

 proceeded

0.89

Activations Density 0.351%

phrases detailing incidents or crimes

No Comments

No Known Activations