INDEX

Explanations

geographical locations and countries

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

ioned

-0.63

xit

-0.61

bp

-0.60

MOT

-0.59

rious

-0.58

hement

-0.57

letes

-0.57

+=

-0.57

zes

-0.57

escription

-0.56

POSITIVE LOGITS

 periphery

0.86

 vicinity

0.84

 where

0.84

 classrooms

0.83

 countryside

0.82

 airspace

0.77

 alone

0.73

 circles

0.73

abouts

0.73

 today

0.73

Activations Density 2.825%

geographical locations and countries

No Comments

No Known Activations