INDEX

Explanations

terms related to graphs or graphical representations

oai_token-act-pair · gpt-3.5-turbo

references to graphs and graphical representations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

mediated

-0.77

artney

-0.71

uders

-0.67

 presumptive

-0.66

akening

-0.63

alle

-0.63

 sworn

-0.61

irmed

-0.61

 Liberties

-0.61

 Kitt

-0.61

POSITIVE LOGITS

 graph

1.16

ically

1.02

 graphs

1.01

graph

0.98

alogy

0.95

 Graph

0.88

 resil

0.88

Graph

0.84

ql

0.83

ļéĨĴ

0.83

Activations Density 0.007%

terms related to graphs or graphical representations

references to graphs and graphical representations

No Comments

No Known Activations