INDEX

Explanations

phrases related to activities or services

oai_token-act-pair · gpt-3.5-turbo

phrases related to governmental activities and operations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 sidx

-0.87

isSpecialOrderable

-0.86

dds

-0.72

 Awakens

-0.72

ENTS

-0.69

ATURE

-0.69

 Barker

-0.67

backer

-0.64

peria

-0.64

 Warn

-0.64

POSITIVE LOGITS

 regulated

1.23

 polic

1.17

 governed

1.14

 segregated

1.07

 handled

1.02

 mediated

1.01

 constrained

1.00

 dictated

0.99

 judged

0.98

 tailored

0.98

Activations Density 0.268%

phrases related to activities or services

phrases related to governmental activities and operations

No Comments

No Known Activations