INDEX

Explanations

technical terms related to technology, hardware, and software

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

natureconservancy

-0.72

Demand

-0.65

untarily

-0.64

hoe

-0.62

"},"

-0.61

atively

-0.61

 guiActiveUnfocused

-0.60

owed

-0.60

sbm

-0.59

ãĤ¼ãĤ¦ãĤ¹

-0.59

POSITIVE LOGITS

 congr

1.34

 beware

1.25

 congratulations

1.04

 please

1.00

 Beware

0.99

 PLEASE

0.96

 remember

0.91

 hello

0.89

 THANK

0.87

 check

0.86

Activations Density 0.159%

technical terms related to technology, hardware, and software

No Comments

No Known Activations