INDEX

Explanations

locations and specific dates within longer written passages

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

emale

-0.66

0000000000000000

-0.66

 vulner

-0.66

llor

-0.65

APD

-0.65

 fortun

-0.64

#$

-0.63

 withd

-0.63

-0.61

jri

-0.61

POSITIVE LOGITS

clus

0.86

clusions

0.83

 lieu

0.80

 vitro

0.79

 Detail

0.79

][

0.77

activated

0.77

cluded

0.76

 vivo

0.76

CLUS

0.76

Activations Density 0.035%

locations and specific dates within longer written passages

No Comments

No Known Activations