INDEX

Explanations

names of political figures

oai_token-act-pair · gpt-3.5-turbo

references to political leaders and significant public figures

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

Ļ½

-0.84

ItemTracker

-0.75

©¶æ¥µ

-0.73

-0.71

¥µ

-0.70

ãĥĥãĥī

-0.69

Ī

-0.67

poon

-0.67

Need

-0.67

¸

-0.67

POSITIVE LOGITS

 reacts

1.42

 attends

1.35

 celebrates

1.33

 listens

1.27

 poses

1.24

 greets

1.23

 participates

1.23

 gestures

1.22

 speaks

1.19

 prepares

1.17

Activations Density 0.237%

names of political figures

references to political leaders and significant public figures

No Comments

No Known Activations