INDEX

Explanations

mentions of specific locations and events involving protests or news conferences

oai_token-act-pair · gpt-3.5-turbo

significant names and locations associated with events

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 Notting

-0.94

London

-0.86

 subway

-0.85

Toronto

-0.82

MTA

-0.78

TTC

-0.78

 Blasio

-0.76

 UEFA

-0.76

ijn

-0.75

adiq

-0.74

POSITIVE LOGITS

 Navajo

1.51

 Idaho

1.50

 Wyoming

1.48

 Yellowstone

1.48

BLM

1.43

 Montana

1.43

 ranch

1.30

 Utah

1.27

 Canyon

1.27

 Nevada

1.26

Activations Density 0.672%

mentions of specific locations and events involving protests or news conferences

significant names and locations associated with events

No Comments

No Known Activations