INDEX

Explanations

details related to war statistics and campaign expenses

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ĸļ

-0.64

 Univers

-0.63

 tremend

-0.62

 bailed

-0.58

 boun

-0.58

 Mara

-0.55

 liber

-0.55

 happ

-0.55

 Hawaiian

-0.55

 stren

-0.54

POSITIVE LOGITS

0.77

0.76

ax

0.75

ollo

0.75

ophers

0.74

0.73

Activations Density 0.027%

details related to war statistics and campaign expenses

No Comments

No Known Activations