INDEX

Explanations

technical terms related to computation

oai_token-act-pair · gpt-3.5-turbo

references to functional relationships or mathematical concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

VIDEO

-0.77

 paycheck

-0.75

funding

-0.69

 dads

-0.69

 vetting

-0.69

Kids

-0.68

 bible

-0.68

south

-0.68

 grassroots

-0.67

 shaming

-0.67

POSITIVE LOGITS

 interpol

1.02

 intermediate

0.94

 synt

0.91

 ÎĶ

0.85

icient

0.85

 analogous

0.84

 characteristic

0.82

 facult

0.82

 âĪ¼

0.81

 compositions

0.81

Activations Density 1.129%

technical terms related to computation

references to functional relationships or mathematical concepts

No Comments

No Known Activations