INDEX

Explanations

phrases related to cleaning up or addressing problems

oai_token-act-pair · gpt-3.5-turbo

references to cleaning or improving situations or environments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

yip

-0.84

isSpecialOrderable

-0.77

thood

-0.72

ï¸

-0.70

yi

-0.69

ESA

-0.69

cair

-0.68

atars

-0.67

zai

-0.65

bos

-0.64

POSITIVE LOGITS

 polluted

0.88

 residue

0.86

 toxins

0.82

 mess

0.80

 backlog

0.78

 cache

0.78

 traces

0.75

 spills

0.75

 contaminated

0.74

 debris

0.73

Activations Density 0.158%

phrases related to cleaning up or addressing problems

references to cleaning or improving situations or environments

No Comments

No Known Activations