INDEX

Explanations

data-related terms or references

oai_token-act-pair · gpt-3.5-turbo

references to data and statistical information in the document

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

ModLoader

-0.91

ovie

-0.80

lain

-0.69

 Oath

-0.67

tsky

-0.66

ength

-0.64

vow

-0.63

complex

-0.63

awar

-0.61

aughs

-0.61

POSITIVE LOGITS

 analytics

0.99

iologist

0.90

 collection

0.85

 recorder

0.85

 data

0.80

 disag

0.80

 collected

0.80

flow

0.77

ãĤ¼ãĤ¦ãĤ¹

0.77

 collectors

0.76

Activations Density 0.033%

data-related terms or references

references to data and statistical information in the document

No Comments

No Known Activations