INDEX

Explanations

comparative adjectives related to quality

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

andise

-0.94

ervative

-0.73

aint

-0.68

ADRA

-0.68

inosaur

-0.65

aukee

-0.63

CLE

-0.63

SU

-0.62

OGR

-0.61

itutional

-0.60

POSITIVE LOGITS

(<

0.98

ered

0.85

enthal

0.82

downs

0.79

est

0.79

case

0.78

down

0.76

 trickle

0.75

 observable

0.75

ppy

0.74

Activations Density 0.686%

comparative adjectives related to quality

No Comments

No Known Activations