INDEX

Explanations

mentions of potential risks or causes related to environmental issues

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

odan

-0.73

ritical

-0.62

reb

-0.61

ĪĴ

-0.61

cemic

-0.61

robe

-0.60

claimer

-0.60

tan

-0.58

ãĥ³ãĤ¸

-0.56

dit

-0.55

POSITIVE LOGITS

 everywhere

1.02

 concurrently

1.01

 elsewhere

0.95

 throughout

0.95

 anywhere

0.93

 across

0.85

 during

0.84

 intermitt

0.83

 here

0.83

 alongside

0.81

Activations Density 4.774%

mentions of potential risks or causes related to environmental issues

No Comments

No Known Activations