INDEX

Explanations

references to unauthorized or third-party software

oai_token-act-pair · gpt-3.5-turbo

references to unauthorized third-party involvement or services

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

veyard

-0.78

è¦ļéĨĴ

-0.68

 ç¥ŀ

-0.68

 Canaver

-0.67

++++++++++++++++

-0.67

 Sturgeon

-0.67

======

-0.66

 Slay

-0.66

veland

-0.65

Reward

-0.64

POSITIVE LOGITS

party

1.12

Party

0.99

 parties

0.93

assian

0.90

 baseman

0.88

ages

0.83

 party

0.78

DN

0.77

member

0.75

part

0.74

Activations Density 0.054%

references to unauthorized or third-party software

references to unauthorized third-party involvement or services

No Comments

No Known Activations