INDEX

Explanations

words related to government and politics

oai_token-act-pair · gpt-3.5-turbo

references to governmental or organizational entities and their initiatives

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

 Luffy

-0.80

 Guerrero

-0.80

SAN

-0.78

 Haku

-0.77

 Albuquerque

-0.77

ppo

-0.76

 Goku

-0.76

 Yamato

-0.76

bors

-0.75

zona

-0.75

POSITIVE LOGITS

 Labour

1.17

 centre

1.16

 recognise

1.10

 councillors

1.10

 scept

1.10

 Ukip

1.09

MPs

1.08

 fibre

1.07

 mould

1.06

 defences

1.05

Activations Density 0.610%

words related to government and politics

references to governmental or organizational entities and their initiatives

No Comments

No Known Activations