INDEX

Explanations

names of locations and institutions

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

ĸļ

-0.91

DERR

-0.76

CLASSIFIED

-0.74

 McDonnell

-0.74

bourg

-0.73

ABE

-0.73

éĹĺ

-0.71

tenance

-0.69

 Commonwealth

-0.68

å£«

-0.67

POSITIVE LOGITS

uana

1.19

Ig

0.99

uala

0.93

omez

0.92

reg

0.88

agi

0.86

les

0.86

uno

0.85

Nob

0.83

min

0.83

Activations Density 0.006%

names of locations and institutions

No Comments

No Known Activations