INDEX

Explanations

organizations or entities related to the British government or British associations

oai_token-act-pair · gpt-3.5-turbo

references to British organizations, institutions, and governmental departments

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

ollah

-0.93

©¶æ

-0.76

xious

-0.75

Houston

-0.71

lando

-0.71

eah

-0.69

ocious

-0.68

php

-0.68

paio

-0.67

 needles

-0.65

POSITIVE LOGITS

 recognised

0.86

 Trave

0.85

 realise

0.83

 Wales

0.82

 Defence

0.81

 Edinburgh

0.80

 Abbey

0.79

 realised

0.77

 recognise

0.77

 Nato

0.75

Activations Density 0.241%

organizations or entities related to the British government or British associations

references to British organizations, institutions, and governmental departments

No Comments

No Known Activations