INDEX

Explanations

people's names with some connection to current or political 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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

eners

-0.69

¬¼

-0.69

HAEL

-0.64

ĸļ

-0.61

creen

-0.61

 comma

-0.60

 birth

-0.59

 blaming

-0.59

 sterling

-0.58

 compounded

-0.57

POSITIVE LOGITS

ansas

1.21

thritis

1.10

ithmetic

0.99

beit

0.98

rang

0.98

duino

0.96

rival

0.96

agon

0.95

ranging

0.94

gent

0.93

Activations Density 0.497%

people's names with some connection to current or political events

No Comments

No Known Activations