INDEX

Explanations

Keywords related to different types of strains

oai_token-act-pair · gpt-3.5-turbo

references to various types of biological strains or pressures

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

*/(

-1.10

emouth

-0.91

fman

-0.90

naissance

-0.88

merce

-0.85

nect

-0.84

orgetown

-0.82

tarians

-0.80

theless

-0.79

issance

-0.78

POSITIVE LOGITS

 strain

1.08

 strains

1.08

 spores

0.81

 bred

0.73

onest

0.69

weed

0.67

igo

0.67

bands

0.65

0.64

 infections

0.63

Activations Density 0.013%

Keywords related to different types of strains

references to various types of biological strains or pressures

No Comments

No Known Activations