INDEX

Explanations

lines of code or technical computer-related terms

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

 breed

-0.79

 venture

-0.75

 resume

-0.74

 limp

-0.71

 coat

-0.70

 pointing

-0.68

 hoard

-0.67

 exchange

-0.66

 releg

-0.66

 absentee

-0.66

POSITIVE LOGITS

ãĥīãĥ©ãĤ´ãĥ³

1.19

true

1.13

PsyNetMessage

1.06

ãĤ¨ãĥ«

1.05

false

1.04

ãĥ´ãĤ¡

1.03

False

1.02

True

1.01

{{

1.01

MpServer

1.00

Activations Density 0.014%

lines of code or technical computer-related terms

No Comments

No Known Activations