INDEX

Explanations

mentions of specific people or political figures

oai_token-act-pair · gpt-3.5-turbo

punctuation and names associated with people

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

vm

-1.01

hov

-0.97

 Robo

-0.96

 droid

-0.95

 robot

-0.90

 Docker

-0.89

 Sebast

-0.84

 fasc

-0.83

 Unreal

-0.83

 pods

-0.82

POSITIVE LOGITS

 Collins

2.57

Collins

2.51

 Sullivan

1.43

Sullivan

1.32

 Maine

1.23

 Gardner

1.21

 Somers

1.19

 Crowley

1.13

 Browne

1.08

 Spect

1.05

Activations Density 0.243%

mentions of specific people or political figures

punctuation and names associated with people

No Comments

No Known Activations