INDEX

Explanations

phrases related to news headlines and political events

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

alties

-0.78

aneers

-0.72

CVE

-0.71

ange

-0.70

rued

-0.67

oi

-0.67

esm

-0.66

oken

-0.66

existent

-0.66

anes

-0.65

POSITIVE LOGITS

 excerpt

0.96

 rundown

0.95

 excerpts

0.84

 snippet

0.81

 highlights

0.78

 overview

0.77

 hoping

0.76

 summary

0.76

 nutshell

0.72

 links

0.71

Activations Density 0.671%

phrases related to news headlines and political events

No Comments

No Known Activations