INDEX

Explanations

titles of TV shows and movies

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

undai

-0.76

 tremend

-0.76

ivable

-0.69

uggest

-0.69

etooth

-0.68

 trouble

-0.67

bered

-0.66

avorite

-0.65

agonal

-0.63

azo

-0.63

POSITIVE LOGITS

 Provided

0.95

 Become

0.85

 https

0.83

 Prelude

0.78

YES

0.78

 http

0.76

hover

0.76

TBD

0.75

 Rise

0.74

 Volume

0.74

Activations Density 0.337%

titles of TV shows and movies

No Comments

No Known Activations