INDEX

Explanations

specific instances of someone telling something

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

 ILCS

-0.82

Interstitial

-0.77

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.73

aband

-0.71

 sidx

-0.70

isol

-0.70

adesh

-0.67

mun

-0.67

 squared

-0.64

ample

-0.64

POSITIVE LOGITS

 reporters

1.14

 BuzzFeed

0.98

 HuffPost

0.94

us

0.94

me

0.91

 Politico

0.85

 Newsweek

0.85

 Fairfax

0.84

 Guardian

0.84

 VICE

0.82

Activations Density 0.060%

specific instances of someone telling something

No Comments

No Known Activations

specific instances of someone telling something

No Comments

No Known Activations