INDEX

Explanations

locations

oai_token-act-pair · gpt-3.5-turbo

names of towns, cities, or geographical locations

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

 Reincarn

-0.72

BILITIES

-0.69

 copied

-0.68

 crank

-0.67

 personality

-0.67

OPLE

-0.67

 subscript

-0.66

 biography

-0.65

 encyclopedia

-0.65

 omission

-0.64

POSITIVE LOGITS

 Harbour

0.97

hend

0.94

ona

0.91

hurst

0.88

istar

0.87

opolis

0.87

dale

0.85

 Avenue

0.84

 Village

0.84

alia

0.84

Activations Density 0.344%

locations

names of towns, cities, or geographical locations

No Comments

No Known Activations