INDEX

Explanations

words related to difficulty or intensity

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

hig

-0.77

Interstitial

-0.69

umbn

-0.66

edy

-0.66

nikov

-0.65

ATOR

-0.64

iae

-0.63

ioxide

-0.63

ãĤ´ãĥ³

-0.63

comings

-0.62

POSITIVE LOGITS

 they

0.95

soever

0.84

we

0.83

he

0.81

 THEY

0.81

you

0.77

she

0.75

 constitutes

0.73

 thou

0.73

 awaits

0.72

Activations Density 0.926%

words related to difficulty or intensity

No Comments

No Known Activations