INDEX

Explanations

phrases or concepts related to legal reasoning

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

chu

-0.81

chin

-0.78

rael

-0.78

CHAT

-0.74

yang

-0.72

tha

-0.71

 Reincarnated

-0.71

osa

-0.69

ocket

-0.68

uria

-0.68

POSITIVE LOGITS

 expectation

1.02

 expectations

0.99

 priced

0.88

 accommodation

0.87

 sized

0.85

 precaution

0.83

 foreseeable

0.82

 accommodations

0.82

 enough

0.82

 amounts

0.81

Activations Density 0.027%

phrases or concepts related to legal reasoning

No Comments

No Known Activations