INDEX

Explanations

references to specific universities

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names of universities and geographical locations

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

 wagon

-0.70

 Tornado

-0.69

 Directive

-0.61

cra

-0.59

 Latest

-0.58

 loyal

-0.58

 FIGHT

-0.58

DPR

-0.57

oshenko

-0.57

ndum

-0.57

POSITIVE LOGITS

University

1.26

 universities

1.17

 University

1.09

College

1.05

orgetown

1.04

 Harvard

1.03

 Faculty

1.02

 Universities

1.00

negie

0.97

 School

0.95

Activations Density 0.217%

references to specific universities

proper nouns, particularly names of universities and geographical locations

No Comments

No Known Activations