INDEX

Explanations

steps or events that need to happen before a certain action

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

aez

-0.74

geist

-0.72

atic

-0.68

OTOS

-0.67

aren

-0.66

SA

-0.66

ophe

-0.64

ingle

-0.64

JV

-0.63

Force

-0.63

POSITIVE LOGITS

 embark

0.92

 committing

0.92

 sunrise

0.89

 proceeding

0.89

 settling

0.87

 heading

0.85

 realizing

0.84

hand

0.83

 entering

0.83

 noon

0.82

Activations Density 0.313%

steps or events that need to happen before a certain action

No Comments

No Known Activations