INDEX

Explanations

locations

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

pter

-0.88

itual

-0.78

alez

-0.78

gio

-0.76

ensitive

-0.75

ensitivity

-0.75

impl

-0.74

ulence

-0.73

byn

-0.72

agonists

-0.71

POSITIVE LOGITS

 lobster

0.83

 Ridge

0.79

 Township

0.78

town

0.75

 Legislature

0.73

 coast

0.71

 Coast

0.70

 Shore

0.69

 Deer

0.69

 Lamp

0.68

Activations Density 0.014%

locations

No Comments

No Known Activations