INDEX

Explanations

phrases indicating lack of connection or relevance to a specific topic

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

©¶æ¥µ

-0.79

DL

-0.76

psons

-0.72

aido

-0.71

kai

-0.71

 Ahead

-0.70

nas

-0.70

hiba

-0.69

uvian

-0.69

pdf

-0.69

POSITIVE LOGITS

 determining

0.87

 upholding

0.86

 politics

0.86

 criminality

0.84

 aesthetics

0.81

 legality

0.81

 deciding

0.80

 ethnicity

0.80

 realism

0.80

 moderation

0.80

Activations Density 0.047%

phrases indicating lack of connection or relevance to a specific topic

No Comments

No Known Activations