INDEX

Explanations

phrases related to espionage and secret agents, specifically focusing on the term "Agent" followed by a number

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

lihood

-1.00

issance

-0.98

etooth

-0.79

FACE

-0.74

ths

-0.73

ĸļ

-0.70

vironment

-0.68

brance

-0.66

phrine

-0.66

amping

-0.65

POSITIVE LOGITS

 prov

1.01

 agent

0.93

agent

0.91

 Agent

0.91

 agents

0.87

agents

0.83

 Agents

0.83

eers

0.81

Agent

0.78

iola

0.78

Activations Density 0.028%

phrases related to espionage and secret agents, specifically focusing on the term "Agent" followed by a number

No Comments

No Known Activations