INDEX

Explanations

personal narratives and discussions about life experiences

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

icol

-0.80

OGR

-0.76

nin

-0.76

mire

-0.75

Written

-0.73

\/\/

-0.72

Ò

-0.71

horn

-0.71

agues

-0.69

warn

-0.69

POSITIVE LOGITS

 regards

1.21

 legality

1.08

 finances

1.05

 aesthetics

1.04

 determining

1.04

 matters

1.02

 regard

1.01

 accessibility

1.00

 whether

1.00

 hygiene

0.98

Activations Density 1.527%

personal narratives and discussions about life experiences

No Comments

No Known Activations