INDEX

Explanations

instances of the word "vote."

oai_token-act-pair · gpt-3.5-turbo

instances of voting-related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 millenn

-0.78

Kut

-0.74

 Palestin

-0.73

 unden

-0.71

 princ

-0.71

Sah

-0.67

Mub

-0.65

cyclopedia

-0.65

gotten

-0.64

Esc

-0.63

POSITIVE LOGITS

 vote

0.93

vote

0.89

board

0.88

 votes

0.87

 unanimously

0.86

room

0.84

imony

0.82

boarding

0.79

rooms

0.77

 Vote

0.75

Activations Density 0.027%

instances of the word "vote."

instances of voting-related terms

No Comments

No Known Activations