INDEX

Explanations

references to legal pardons and specific locations like "Romeo."

oai_token-act-pair · gpt-3.5-turbo

words related to "pardon" and its various forms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

oted

-0.71

pled

-0.71

elled

-0.69

urg

-0.67

 parcel

-0.66

 auditor

-0.64

pling

-0.62

rus

-0.62

 uninsured

-0.62

 liter

-0.61

POSITIVE LOGITS

ardon

1.28

ably

0.93

MENTS

0.81

åī

0.79

abouts

0.77

 pardon

0.75

 pard

0.75

itial

0.72

 "$:/

0.72

rontal

0.71

Activations Density 0.015%

references to legal pardons and specific locations like "Romeo."

words related to "pardon" and its various forms

No Comments

No Known Activations