INDEX

Explanations

words related to sports teams and their performances

oai_token-act-pair · gpt-3.5-turbo

references to different teams or groups in a competitive context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

jri

-0.89

ESCO

-0.76

erous

-0.72

gently

-0.71

MFT

-0.69

igslist

-0.69

ertodd

-0.68

lus

-0.67

zie

-0.67

INESS

-0.67

POSITIVE LOGITS

kick

0.89

wich

0.81

por

0.73

plan

0.72

Ole

0.72

footed

0.72

 Marse

0.71

AFC

0.69

Sax

0.69

FC

0.67

Activations Density 0.019%

words related to sports teams and their performances

references to different teams or groups in a competitive context

No Comments

No Known Activations