INDEX

Explanations

locations or places, especially emphasized in text related to various objects or events

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

vous

-0.86

jin

-0.79

missions

-0.75

unes

-0.74

tails

-0.74

sonian

-0.73

squ

-0.71

xtap

-0.70

ework

-0.69

sle

-0.69

POSITIVE LOGITS

 location

1.00

 locations

0.98

 locate

0.79

 locality

0.76

 nearest

0.75

 scout

0.75

 vicinity

0.72

 where

0.71

URA

0.70

 Location

0.69

Activations Density 0.051%

locations or places, especially emphasized in text related to various objects or events

No Comments

No Known Activations