INDEX

Explanations

explanations or reasoning in a text

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

icipated

-0.66

apers

-0.66

rift

-0.64

iership

-0.63

RAW

-0.63

actionDate

-0.62

shaw

-0.61

vez

-0.60

display

-0.60

ourse

-0.58

POSITIVE LOGITS

 yeah

1.04

yeah

1.04

hhh

0.99

hhhh

0.95

 prest

0.94

 Yeah

0.93

 kidding

0.92

 pardon

0.89

mmm

0.88

yea

0.87

Activations Density 0.648%

explanations or reasoning in a text

No Comments

No Known Activations