INDEX

Explanations

references to electronic systems and specific terms related to access and security

oai_token-act-pair · gpt-3.5-turbo

references to electronic systems and their security

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Beir

-0.94

psons

-0.91

 Antar

-0.91

EPA

-0.87

bert

-0.78

anto

-0.77

reau

-0.75

ãĥ´ãĤ¡

-0.75

 Fran

-0.71

 piping

-0.71

POSITIVE LOGITS

 lock

2.20

 Lock

2.19

Lock

2.08

 locks

2.03

lock

1.90

 locking

1.82

locks

1.77

LOCK

1.69

 locked

1.64

 Locked

1.64

Activations Density 0.269%

references to electronic systems and specific terms related to access and security

references to electronic systems and their security

No Comments

No Known Activations