INDEX

Explanations

phrases related to requests or commands

oai_token-act-pair · gpt-3.5-turbo

phrases indicating requests or demands

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

natureconservancy

-0.64

 sear

-0.63

 testament

-0.60

aneous

-0.59

Zip

-0.58

 <!--

-0.57

ua

-0.55

Cause

-0.55

 Appalach

-0.54

ect

-0.54

POSITIVE LOGITS

ħĭ

0.84

otom

0.82

isson

0.81

pez

0.78

by

0.78

ãĥīãĥ©

0.77

by

0.76

iak

0.73

ocobo

0.70

å§«

0.70

Activations Density 0.360%

phrases related to requests or commands

phrases indicating requests or demands

No Comments

No Known Activations