INDEX

Explanations

second-person pronouns

oai_token-act-pair · gpt-3.5-turbo

references to personal pronouns indicating involvement or action

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Rhodes

-0.74

 Pegasus

-0.70

JFK

-0.69

 Forth

-0.69

 Newark

-0.68

 Kissinger

-0.67

 Liberty

-0.67

Gad

-0.66

 Atlantis

-0.62

 Keynes

-0.60

POSITIVE LOGITS

âĢ

2.12

 âĢ

1.46

Â«

1.39

âĶ

1.38

Â»

1.36

ãĢ

1.34

Â¨

1.30

ÃĥÃĤ

1.29

Ò

1.27

îĢ

1.26

Activations Density 0.356%

second-person pronouns

references to personal pronouns indicating involvement or action

No Comments

No Known Activations