INDEX

Explanations

terms related to risk analysis and management

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

gdala

-0.83

atters

-0.72

issance

-0.68

eve

-0.64

Writer

-0.63

ADRA

-0.63

elf

-0.63

riot

-0.63

ergy

-0.62

Bed

-0.62

POSITIVE LOGITS

 posed

0.96

 hazards

0.92

 incurred

0.88

 factors

0.88

 mitigation

0.88

 aversion

0.87

 risks

0.87

 avoidance

0.83

 advis

0.81

taking

0.80

Activations Density 0.555%

terms related to risk analysis and management

No Comments

No Known Activations