INDEX

Explanations

terms related to emotional and moral concepts of belief and intention

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

Thirty

-0.89

ULAR

-0.79

USS

-0.75

YR

-0.74

Recomm

-0.73

ãĥĥãĥī

-0.71

Specific

-0.70

URA

-0.67

Stars

-0.66

trak

-0.66

POSITIVE LOGITS

edly

1.09

 syndrome

0.86

nered

0.85

 affair

0.84

 prevailed

0.75

ned

0.74

 attitude

0.70

 impression

0.70

 excuse

0.68

 prevail

0.67

Activations Density 0.050%

terms related to emotional and moral concepts of belief and intention

No Comments

No Known Activations