INDEX

Explanations

references to specific countries, in this case, particularly India

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

icipated

-0.80

iliary

-0.75

checks

-0.74

tower

-0.73

esters

-0.73

rd

-0.73

mble

-0.71

RH

-0.69

cer

-0.69

prus

-0.68

POSITIVE LOGITS

apolis

0.97

 Sharma

0.94

Rao

0.85

ru

0.84

 Kumar

0.82

 Gupta

0.82

endra

0.81

 Pradesh

0.81

 Punjab

0.78

 Singh

0.78

Activations Density 0.931%

references to specific countries, in this case, particularly India

No Comments

No Known Activations