INDEX

Explanations

phrases related to security measures and potential threats

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

blance

-0.68

dayName

-0.67

ertodd

-0.63

 redesign

-0.61

 disposed

-0.59

overe

-0.59

 condem

-0.57

OGR

-0.57

pots

-0.56

ritic

-0.56

POSITIVE LOGITS

us

0.89

 these

0.76

 them

0.73

them

0.73

 course

0.71

these

0.69

course

0.69

 sudden

0.68

 those

0.67

our

0.66

Activations Density 0.406%

phrases related to security measures and potential threats

No Comments

No Known Activations