INDEX

Explanations

historical and medical-related text

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

-0.75

driving

-0.73

Interstitial

-0.66

 imagining

-0.65

ignt

-0.64

ragon

-0.62

ovan

-0.61

instead

-0.61

thouse

-0.60

eday

-0.60

POSITIVE LOGITS

 Overview

1.17

 Edit

1.17

 Description

1.17

 References

1.15

 Notes

1.15

 Summary

1.15

Â¶

1.13

 Definitions

1.06

 Contents

1.06

 Supported

1.04

Activations Density 1.386%

historical and medical-related text

No Comments

No Known Activations