INDEX

Explanations

the word "Chan" with varying degrees of activation, potentially indicating a focus on a specific person or concept named Chan

oai_token-act-pair · gpt-3.5-turbo

references to the term "Chan" 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 @ 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

ACTION

-0.80

ACTED

-0.78

GROUND

-0.77

 mileage

-0.75

OAD

-0.75

ÃįÃį

-0.70

IGH

-0.69

IBLE

-0.69

 Curiosity

-0.68

ãģ¾

-0.66

POSITIVE LOGITS

ning

1.25

bara

1.12

istry

0.96

bach

0.96

ned

0.95

sey

0.91

eless

0.91

unci

0.91

plain

0.85

oir

0.85

Activations Density 0.027%

the word "Chan" with varying degrees of activation, potentially indicating a focus on a specific person or concept named Chan

references to the term "Chan" and its variations

No Comments

No Known Activations