INDEX

Explanations

words related to alternative or unseen realms like "ether" and "Nether"

oai_token-act-pair · gpt-3.5-turbo

mentions of the term "ether" and its variations

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

ļé

-0.69

©¶æ

-0.65

 chef

-0.63

 traumatic

-0.63

EU

-0.62

ushima

-0.62

QUI

-0.61

 Strauss

-0.60

ured

-0.59

Naz

-0.59

POSITIVE LOGITS

sburg

1.14

rack

1.01

spr

0.98

icut

0.97

ionage

0.96

0.95

apy

0.92

ship

0.92

hemy

0.91

ether

0.91

Activations Density 0.020%

words related to alternative or unseen realms like "ether" and "Nether"

mentions of the term "ether" and its variations

No Comments

No Known Activations