INDEX

Explanations

phrases related to contrasts or comparisons

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

OGR

-0.87

unc

-0.84

gow

-0.81

nit

-0.75

vous

-0.74

nell

-0.73

quished

-0.72

bye

-0.69

 teasp

-0.69

TPPStreamerBot

-0.69

POSITIVE LOGITS

 sexes

1.11

 genders

1.00

 halves

0.93

 extremes

0.86

 factions

0.82

 sides

0.80

 combatants

0.79

 them

0.76

two

0.72

 parties

0.71

Activations Density 0.403%

phrases related to contrasts or comparisons

No Comments

No Known Activations