INDEX

Explanations

occurrences of the word "whole."

oai_token-act-pair · gpt-3.5-turbo

instances of the word "whole" related to completeness or entirety

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.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

osponsors

-0.87

ics

-0.68

KER

-0.65

akers

-0.64

ufact

-0.64

orthy

-0.63

nell

-0.61

fu

-0.60

zynski

-0.60

umper

-0.60

POSITIVE LOGITS

heartedly

1.21

hearted

1.04

 thing

0.94

 darn

0.86

 ordeal

0.85

gam

0.80

 damn

0.78

around

0.77

 freaking

0.74

 globe

0.71

Activations Density 0.026%

occurrences of the word "whole."

instances of the word "whole" related to completeness or entirety

No Comments

No Known Activations