INDEX

Explanations

references to the Minnesota Vikings football team

oai_token-act-pair · gpt-3.5-turbo

mentions of the state of Minnesota and its sports teams

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

arding

-0.88

arded

-0.82

ARDS

-0.74

pert

-0.73

ours

-0.71

spr

-0.70

ards

-0.70

bara

-0.69

ulator

-0.68

rawling

-0.68

POSITIVE LOGITS

 Timberwolves

1.21

Minnesota

1.13

 Vikings

1.11

Dul

1.09

 Twins

1.01

nesota

0.99

 Minnesota

0.95

burgh

0.82

Fay

0.81

 Cities

0.79

Activations Density 0.015%

references to the Minnesota Vikings football team

mentions of the state of Minnesota and its sports teams

No Comments

No Known Activations