INDEX

Explanations

words related to geographical locations or affiliations with specific places

oai_token-act-pair · gpt-3.5-turbo

mentions of locations or organizations linked to specific cities or regions

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

 Dickinson

-0.78

 Grimes

-0.71

 Norris

-0.68

cir

-0.65

 shade

-0.65

 Finn

-0.64

 Indigo

-0.63

 Farrell

-0.63

 interval

-0.62

oleon

-0.62

POSITIVE LOGITS

based

1.34

centric

1.20

style

1.19

themed

1.15

esque

1.10

sized

1.09

derived

1.07

driven

1.07

exclusive

1.05

inspired

1.04

Activations Density 0.134%

words related to geographical locations or affiliations with specific places

mentions of locations or organizations linked to specific cities or regions

No Comments

No Known Activations