INDEX

Explanations

terms related to health care, medical treatments, and government healthcare policies

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

luaj

-0.78

ngth

-0.77

opez

-0.70

ividual

-0.66

isively

-0.62

raper

-0.62

aneers

-0.61

asers

-0.59

ifferent

-0.59

elve

-0.58

POSITIVE LOGITS

 happening

0.94

 true

0.81

 untrue

0.80

why

0.77

 blasphemy

0.75

 unacceptable

0.75

 understandable

0.75

SPONSORED

0.71

 heresy

0.71

 nonsense

0.70

Activations Density 2.084%

terms related to health care, medical treatments, and government healthcare policies

No Comments

No Known Activations