INDEX

Explanations

references to the word "Pal" in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

ãĥ¼ãĥĨãĤ£

-0.77

REE

-0.70

hered

-0.68

ICLE

-0.64

ELF

-0.62

 raging

-0.61

 abdom

-0.60

hift

-0.60

net

-0.60

ï¸

-0.59

POSITIVE LOGITS

estine

1.28

atial

1.18

adin

1.17

grave

1.13

myra

1.11

mers

0.99

atin

0.99

igans

0.98

agus

0.97

mares

0.96

Activations Density 0.006%

references to the word "Pal" in various contexts

No Comments

No Known Activations