INDEX

Explanations

references to "micro" and "macro" concepts or technologies

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

UAL

-0.76

aukee

-0.73

swick

-0.69

IUM

-0.69

 Guilty

-0.68

igham

-0.67

IELD

-0.67

atche

-0.66

Eve

-0.65

lain

-0.65

POSITIVE LOGITS

organisms

1.29

gravity

1.25

seconds

1.22

cos

1.12

cephal

1.05

chip

1.04

bial

1.03

processor

1.01

econom

1.01

climate

1.01

Activations Density 0.011%

references to "micro" and "macro" concepts or technologies

No Comments

No Known Activations