INDEX

Explanations

terms related to broader concepts or contexts

oai_token-act-pair · gpt-3.5-turbo

the term "broader," indicating discussions of wider or more inclusive contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 Tycoon

-0.78

LV

-0.74

 Athlet

-0.70

çļ

-0.69

boy

-0.68

ILLE

-0.68

liest

-0.66

 Commando

-0.64

 Steal

-0.63

SR

-0.63

POSITIVE LOGITS

 reflection

0.87

 contextual

0.83

 guiActiveUn

0.83

ynski

0.83

unci

0.81

sheets

0.81

soType

0.81

izont

0.81

 behavi

0.80

 societal

0.80

Activations Density 0.007%

terms related to broader concepts or contexts

the term "broader," indicating discussions of wider or more inclusive contexts

No Comments

No Known Activations