INDEX

Explanations

What the neuron is looking for: movements or actions related to physical activities

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

saf

-0.84

laus

-0.69

 narrower

-0.67

pai

-0.65

©¶æ

-0.65

wich

-0.65

($)

-0.65

ollow

-0.64

lict

-0.64

neau

-0.64

POSITIVE LOGITS

 adrenaline

1.20

 excitement

1.15

 anticipation

0.99

 frenzy

0.99

 fury

0.96

 ecstasy

0.95

 enthusiasm

0.94

 adren

0.93

 pandemonium

0.93

 roaring

0.90

Activations Density 0.259%

What the neuron is looking for: movements or actions related to physical activities

No Comments

No Known Activations