INDEX

Explanations

statements of admission or confession

oai_token-act-pair · gpt-3.5-turbo

instances of the word "admitting" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Commando

-0.64

 Properties

-0.63

Nim

-0.60

awar

-0.59

orp

-0.59

 Rally

-0.59

urry

-0.58

rik

-0.58

 tick

-0.58

Tim

-0.57

POSITIVE LOGITS

 admitting

3.91

 admit

2.24

 admits

2.03

 admitted

1.88

 acknowledging

1.80

 conced

1.78

 admission

1.67

 confess

1.63

 apologizing

1.37

 concedes

1.37

Activations Density 0.014%

statements of admission or confession

instances of the word "admitting" and its variations

No Comments

No Known Activations