INDEX

Explanations

words and phrases related to negative connotations and judgements

oai_token-act-pair · gpt-4o-mini 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

 shape

-0.81

 warr

-0.78

 ende

-0.75

 nude

-0.74

 barg

-0.72

 dreamed

-0.72

 imagination

-0.72

 jugg

-0.71

 untouched

-0.71

 unborn

-0.71

POSITIVE LOGITS

ï¸ı

1.10

 Additionally

1.10

 Similarly

1.10

However

1.08

 Unfortunately

1.08

 However

1.06

Additionally

1.03

 Likewise

1.03

ï¸

1.01

 Therefore

1.01

Activations Density 0.148%

words and phrases related to negative connotations and judgements

No Comments

No Known Activations