INDEX

Explanations

descriptions and sentiments related to experiences or qualities

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

dit

-0.81

bernatorial

-0.74

bus

-0.71

KO

-0.71

otti

-0.70

orthy

-0.70

umi

-0.69

byter

-0.69

swick

-0.67

thia

-0.66

POSITIVE LOGITS

 these

0.99

 this

0.98

 owning

0.85

 THESE

0.78

CoC

0.77

 storytelling

0.73

our

0.73

these

0.72

 constructing

0.71

 life

0.71

Activations Density 1.708%

descriptions and sentiments related to experiences or qualities

No Comments

No Known Activations