INDEX

Explanations

words related to a specific person's name 'Fram' combined with numbers or other text phrases

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

cules

-0.70

IRO

-0.68

 DRAG

-0.65

wana

-0.63

 perm

-0.62

Merit

-0.61

icably

-0.60

 motors

-0.60

 Bulg

-0.60

âĵĺ

-0.59

POSITIVE LOGITS

eworks

1.24

erate

1.05

erity

1.03

pton

1.00

etimes

0.98

selage

0.96

estine

0.95

ework

0.91

eline

0.89

etime

0.86

Activations Density 0.029%

words related to a specific person's name 'Fram' combined with numbers or other text phrases

No Comments

No Known Activations