INDEX

Explanations

numerical data points or statistics related to various topics

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

akening

-0.78

atism

-0.76

 Vaugh

-0.75

urrent

-0.74

ĸļ

-0.73

rolet

-0.71

mble

-0.68

ricular

-0.67

fare

-0.67

ategory

-0.67

POSITIVE LOGITS

1.20

rd

1.02

LECT

0.88

mm

0.79

aan

0.77

RD

0.76

RF

0.75

inately

0.74

AW

0.74

Â¯Â¯Â¯Â¯

0.72

Activations Density 0.031%

numerical data points or statistics related to various topics

No Comments

No Known Activations