INDEX

Explanations

technical terms related to data analysis and settings

oai_token-act-pair · gpt-3.5-turbo

references to parameters in a technical or scientific context

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ILLE

-0.81

BY

-0.78

bery

-0.72

Es

-0.72

minster

-0.69

 coli

-0.68

IDA

-0.67

Ķ

-0.65

ALK

-0.64

Poe

-0.62

POSITIVE LOGITS

 parameters

1.32

ilitary

1.16

eters

1.01

 parameter

0.95

userc

0.95

eter

0.92

 Parameters

0.91

 flares

0.88

ftime

0.86

etric

0.84

Activations Density 0.006%

technical terms related to data analysis and settings

references to parameters in a technical or scientific context

No Comments

No Known Activations